How Can Leaders Overcome the Blurry Vision Bias ......BRIAN J. LUCAS Cornell University Evidence...

r Academy of Management Journal2018, Vol. 61, No. 6, 2106–2129.https://doi.org/10.5465/amj.2015.0375

HOW CAN LEADERS OVERCOME THE BLURRY VISIONBIAS? IDENTIFYING AN ANTIDOTE TO THE PARADOX

OF VISION COMMUNICATION

ANDREW M. CARTONUniversity of Pennsylvania

BRIAN J. LUCASCornell University

Evidence suggests that organizational leaders can inspire employees by communicatinga vision of the future with image-based rhetoric—words and phrases that are readilyenvisioned in the mind’s eye (e.g., “our vision is to make moviegoers laugh”). Yet researchhas demonstrated that most leaders do not craft visions with image-based rhetoric, insteadfavoring abstract language that cannot easily be visualized. We integrate theory on lead-ership and dual cognitive processing to argue that this problem is exacerbated whenleaders focus on word selection when crafting visions because they overemphasize themeaning-based cognitive system (in which they consider the abstract meaning of words)and underemphasize the experience-based cognitive system (in which they can generatevividmental images of what the future could look like). We introduce a novel tactic to helpleaders activate the experience-based system and, in turn, generate and communicatemore impactful visions. We also investigate boundary conditions. We test our predictionswith three experiments featuring three distinct samples, including one with senior cor-porate executives and one in which members of the British government crafted visions onthe day Britain announced it would exit the European Union (“Brexit”).

Some of the most successful organizational en-deavors have been credited to leaders who galva-nized collective action by articulating a compellingpurpose. For example, Bill Gates envisioneda “computer on everydesk and in everyhome,”1 JohnF. Kennedy challenged NASA to land “a man on themoon,”2 and Henry Ford imagined a “motor car forthe great multitude” that people could enjoy for“hours of pleasure in God’s great open spaces.”3 The

extent to which leaders effectively articulate a senseof purpose is central to our understanding of not onlyhow they inspire action, but the very concept ofleadership itself, because leading is defined as theprocess of influencing others to achieve a commonpurpose (Locke, 1999). Indeed, one of the few pointsof overlap among a range of modern leadership the-ories is that the ability to convey a sense of purpose isone of themost essential leadership skills (Bass, 1990;Greer, Homan,DeHoogh, &DenHartog, 2012;Nanus,1992; Stam, Lord, van Knippenberg, & Wisse, 2014;van Knippenberg & Sitkin, 2013).

The above examples of Gates, Kennedy, and Fordare telling not only because each leader articulateda compelling purpose, but because they did so byemploying image-based rhetoric, whichwe define aslanguage that depicts objects (e.g., “cars”), actions(e.g., “driving”), and events (e.g., “landing on themoon”) that can be observed with one’s senses.Leaders can use image-based rhetoric to enliven oneof the most important rhetorical tactics for commu-nicating purpose: the vision (Stam et al., 2014).Compared to abstract rhetoric (e.g., “aiming for ex-cellence”), image-based rhetoric engages employees’emotions and provides a shared point of reference

We thankAssociateEditorDaanvanKnippenbergand threeanonymous reviewers for their investment in the paperthroughout the review process. For helpful comments, wethankSreedhariDesai,LanceFerris,SamirNurmohamed,KatyMilkman, Keith Murnighan, Maurice Schweitzer, and col-leagues in seminars at INSEAD,TempleUniversity,Universityof Southern California, University of Virginia, andWharton.

1 The full quote is available at: https://www.telegraph.co.uk/technology/3357701/Bill-Gatess-dream-A-computer-in-every-home.html.

2 The full quote is available at: https://www.jfklibrary.org/Asset-Viewer/Archives/JFKWHA-032.aspx.

3 The full quote is available at: https://www.thehenryford.org/collections-and-research/digital-resources/popular-topics/henry-ford-quotes/.

2106

Copyright of the Academy of Management, all rights reserved. Contents may not be copied, emailed, posted to a listserv, or otherwise transmitted without the copyright holder’s expresswritten permission. Users may print, download, or email articles for individual use only.

https://doi.org/10.5465/amj.2015.0375

https://www.telegraph.co.uk/technology/3357701/Bill-Gatess-dream-A-computer-in-every-home.html



https://www.jfklibrary.org/Asset-Viewer/Archives/JFKWHA-032.aspx

https://www.jfklibrary.org/Asset-Viewer/Archives/JFKWHA-032.aspx

https://www.thehenryford.org/collections-and-research/digital-resources/popular-topics/henry-ford-quotes/



around which different subgroups can coordinate(Emrich, Brower, Feldman, & Garland, 2001). Byvividly depicting an event or outcome that an orga-nization can one day realize, image-based rhetoricreflects the notion that a vision is a “portrait” of anideal future (Rafferty & Griffin, 2004) and under-scores the very essence of the word “vision”—theability to see.

Despite the clear reasons for crafting and com-municating vivid visions, leaders tend to communi-cate visions with abstract rhetoric (Carton, Murphy,&Clark, 2014; Emrich et al., 2001), a phenomenonwecall the “blurry vision bias.” This bias largely stemsfrom the reality that people tend to thinkabstractly asthey ponder the distant future (Trope & Liberman,2003). Given that dominant perspectives on visioncommunication assume that leaders can improvevision communication by carefully attending toword selection (Gardner & Schermerhorn, 2004),the solution to the blurry vision bias is ostensiblystraightforward: leaders should consciously focus onincorporating image-based rhetoric into their vi-sions. Consistent with this idea, those who argue forthe value of imagery in vision communication haveimplicitly assumed that leaders who are informed ofthe importance of this type of rhetoric will be fullycapable of incorporating it into their visions (Cartonet al., 2014; Collins & Porras, 1994; Emrich et al.,2001).

We challenge this assumption by evaluatingtheory on vision communication through thelens of dual cognitive processing. When peoplefocus their attention directly on language, theyactivate the part of their mind—themeaning-basedsystem—responsible for considering abstract con-cepts (“superior customer service”) rather than thepart of the mind—the experience-based system—

responsible for processing sensory informationabout the outside world and mentally simulatingreal life experiences (“watching customers smilingas they dine in our lively restaurants”). Whenleaders do not have an image in theirmind, they areless likely to employ image-based rhetoric in theirvisions. Thus, even leaders who deliberately focuson constructing visions with image-laden rhetoricare likely to continue to craft and communicateabstract visions.

To identify an antidote to the blurry visionbias,wefurther develop the link between leader communi-cation and dual cognitive processing. We examinea tactic—mentally projecting oneself to a momentin the distant future, akin to mental time travel(Suddendorf, Addis, & Corballis, 2009)—that targets

the experience-based system rather than themeaning-based system. Rather than contemplate thedistant future abstractly, this tactic impels leaders toimagine the future in vivid detail, as if they are di-rectly observing a future scenario through firsthandobservation. When leaders who take a mental leapforward in time are prompted to describe theirmental imagery in words, they reflexively employimage-based rhetoric since it represents the mostappropriate type of language for communicatingvivid images. We further harness theory on dualcognitive processing to identify a personality dif-ference (efficient versus analytical thinking) thatexplains which leaders benefit most from imagin-ing how the distant future will be experiencedfirsthand. We test our predictions in threeexperiments—one launched on the day of “Brexit”(June 24, 2016, when Britain announced it wasleaving the European Union) that featured Britishgovernment employees (Study 1); one featuringleaders with various spans of control (Study 2); andone exclusively featuring upper echelon leaders(Study 3). In this latter study we assess whether thevisions executives crafted as a result of our in-tervention inspired their employees.

Our findings redirect theory on leadership in sev-eral ways. Through a framework informed by dualcognitive processing, we shed light on a series of keyquestions pertaining to vision communication: whythe blurry vision bias exists, what tactic can correctit, how this tactic works (mediating processes), andwhen it works most effectively (moderating pro-cesses). In doing so, we first explain why an as-sumption that underlies dominant perspectives intheory on vision communication as well as the in-tuition of leaders themselves (that leaders should fo-cus intently on word selection), is incongruent withthe nature of dual cognitive processing. We then in-troduce a parsimonious framework that brings theoryon vision communication and dual cognitive pro-cessing into greater alignment. As such, we not onlyprovide a deeper understanding of a key leader in-fluence tactic, but also answer calls to more preciselychart the intersections between research on leader-ship, linguistics, and cognition (Shondrick, Dinh,& Lord, 2010). At a broader level, this integratedframework can change our understanding of howleaders should incorporate imagery on other occa-sions when they tend to use overly vague language,such as when they provide feedback and give taskinstructions (Kluger & DeNisi, 1996). We also discussempirical contributions, including evidence for rep-licability via preregistered, high-powered designs

2018 2107Carton and Lucas

(Simmons, Nelson, & Simonsohn, 2011), and evi-dence for various forms of validity. Finally, our ef-fort is practically useful: given that leaders areunlikely to increase their use of image-based rhet-oricmerely by being armedwith the knowledge thatsuch rhetoric is beneficial, our research illuminatesan intervention leaders can use to create a moreinspiring view of what their organizations can oneday achieve.

THE POWER OF IMAGE-BASED RHETORIC INVISION COMMUNICATION

Meeting a person who benefits from one’s help ismore motivating than reading about it (Grant,Campbell, Chen, Cottone, Lapedis, & Lee, 2007);seeing one person die up close affects moral judg-ment more than does dropping a bomb on thou-sands from a distance (Eyal, Liberman, & Trope,2008); and observing a measuring cup containingthe amount of sugar in one soda deters soft drinkconsumption more than reading about caloric con-tent (Heath & Heath, 2010). The effect of this type ofvivid detail is not limited to when people witnessevents first hand, but extends towhen they read textor listen to rhetoric that brings this type of vividdetail to mind (Fletcher, Frith, Baker, Shallice,Frackowiak, & Dolan, 1995). In contrast to wordsthat capture ideas and generalities (“liberty,” “dif-ference,” “excellence”), words that represent theobservable world (e.g., “smile,” “jump,” “yellow”)cause people to construct mental images by draw-ing on a cognitive store of prior encounters withexternal reality. Research on the vividness heuristic(Nisbett & Ross, 1980) and the identifiability heu-ristic (Nordgren & McDonnell, 2011) has demon-strated that verbal descriptions of images triggeremotions more powerfully compared to statisticsand concepts (Bator & Cialdini, 2000). Beyonddriving affective responses, image-based rhetoriccompares favorably to abstract rhetoric because itprovides a clearer sense of direction (Kluger &DeNisi, 1996). The impact of image-laden rhetoricon emotion and clarity causes people to be spurredinto action more than does abstract rhetoric.For instance, a story of a single hungry child elic-ited more charitable giving compared to statisticsabout thousands of starving villagers (Small,Loewenstein, & Slovic, 2007), and people weremore likely to quit smoking when they read abouthow their habit caused people to “reel back in dis-gust from the smoker’s putrid odor” thanwhen they

read about its adverse health effects (Smith &Shaffer, 2000: 777).4

Communicators can use image-laden rhetoricnot only to depict the present, but also to providea snapshot of events that have not yet transpired. Assuch, leaders can employ imagery when they artic-ulate their organizations’ visions of the future. Toillustrate, Ekso (a company that makes robotic suitsthat help paralyzed people walk) created a vision tohelp one million people get up from their wheel-chairs andwalk (Edison Investment Research, 2014).Similar to howpeople find image-based descriptionsof the present more compelling and clarifying thanabstract descriptions of the present, they find image-laden visions of the future more emotionally engag-ing and easier to understand than abstract visionsbecause they can simulate future events in their“mind’s eye” rather than contemplate the future ingeneral terms (Masuda, Kane, Shoptaugh, & Minor,2010). This realism boosts motivation. In one study,registered voters from the U.S. who were given animage-laden vision of the U.S. government on themorning of Election Day 2016 were 11.3% moremotivated to vote compared to participants given anabstract vision, and in another study full-time em-ployees who were given an image-based visionworked for 47.8% longer on a creative task thanemployees who were given an abstract vision.5 Thislatter effect was partially mediated by the vision’semotional impact and ease of comprehension, sug-gesting that imagery is both invigorating and clari-fying. A vivid vision may also boost motivation byhelping employees feel more connected to the orga-nization’s overarching purpose, because employeeswill feel that their day-to-day work is more closelytied to an event that can bewitnessed or experiencedone day in the future than to an end state that is ab-stract, amorphous, and unlikely to transpire in reallife (Carton, 2018). Notably, the impact of image-laden rhetoric on followers extends beyond motiva-tion: an examination of former U.S. President BarackObama’s visionary rhetoric on diversity found that

4 Image-based rhetoric also explains variation in per-formance above and beyond goal specificity (Carton et al.,2014),which is typically conceptualizedas thenarrownessof a range of possible numeric performance targets (Locke& Latham, 1990). This is because numbers, despite beingspecific, are difficult to visualize (Guadagno, Rhoads, &Sagarin, 2011), and thus do not impact emotion and at-tention in the same way that image-laden rhetoric does.

5 We provide information on both studies’ methods inAppendix A at blurryvisionbias.wordpress.com.

2108 DecemberAcademy of Management Journal

http://blurryvisionbias.wordpress.com

his use of image-laden phrases reduced prejudicecompared to when he used abstract rhetoric(Stephens-Davidowitz, 2017).

Beyond boosting individual-level outcomes,image-based rhetoric has benefits for collectives be-cause of its effect on social cognition. Vivid snap-shots of the future, such as a movie studio’s vision to“make moviegoers laugh,” have only one basic in-terpretation, enabling followers from different sub-groups to understand a vision’s “central merits” inthe same way (Stam et al., 2014: 1184–1185). Thishas implications for coordination:whenhundreds orthousands of individuals and subgroups possess notonly amental image of the future, but the samementalimage, they can more easily coalesce to realize a col-lective achievement. In one study, an image-basedvision more strongly established a shared sense ofpurpose than did an abstract vision, and, in turn, theformer helped boost group coordination and perfor-mance (Carton, Murphy, & Clark, 2014).6 Reflectingthese findings, image-laden rhetoric has been creditedfor galvanizing collective action—including duringperiodsof changeanduncertainty,whencoordinationis challenging (Carton, 2018). For instance, MartinLuther King’s “I Have a Dream” speech is thought tohave helped spark the U.S. civil rights movement inpart because of the glimpses King provided of a futurein which “the sons of former slaves and the sons offormer slave owners will be able to sit down together”(Mount, 2010). Similarly, Susan B. Anthony helpedadvance women’s rights by conveying a vision ofa world that invoked women’s equality not only ina general sense, but with language that promptedpeople to visualize women voting and writing lawsalongside men (Stanton, Gordon, & Anthony, 1997).

In addition toenhancing followeroutcomes, leaderswho use image-based words improve their ownstanding. In two samples, Emrich et al. (2001) foundthat U.S. presidents who used image-based rhetoricwere perceived to be more charismatic. Emrich et al.(2001) argued that image-based rhetoric has powerfuleffects on recipients’ emotion and cognition, causingfollowers to view leaders more favorably.

When this evidence is taken together, we workfrom the premise that image-based visions are morebeneficial compared to abstract visions for leadersas well as the individuals and groups they oversee.Moreover, for all their benefits, research has not yetidentified clear downsides to vivid visions. Relativeto abstract visions, image-laden visions do not re-strict decision alternatives or suppress the ability fororganizations to pursue paths that serendipitouslypresent themselves. Consider a vision to “makecustomers smile.” Since this vision can be realizedvia countless products and services, it does notconstrain action.

ILLUMINATING ADESCRIPTIVE–PRESCRIPTIVE GAP

Despite the upside of image-based rhetoric, lessthan 10% of leaders communicate visions withstrong imagery—and, in total, they tend to commu-nicate three to 15 times as much conceptual rhetoricas image-laden rhetoric when articulating visions(Carton et al., 2014; Emrich et al., 2001). Rather thanallowing people to “see” the futurewith rhetoric thatdepicts scenes with graphic, ambient detail aboutobjective reality, most visions almost exclusivelycontain lofty postulations (e.g., “change the world”or “serve the community”) that possess multiple in-terpretations andpush people tomerely consider thefuture. As such, most visions are not truly “vision-ary.” This pattern, which we call the blurry visionbias, is surprising not only because of the benefits ofimage-laden rhetoric but also because imagery mapsonto the historical purpose of a vision, which is tocounter the conceptuality of other rhetorical state-ments, including missions (Baetz & Kenneth, 1998),ideological appeals (Grant & Hofmann, 2011), andstrategic objectives (Miller & Cardinal, 1994).

To understand how the blurry vision bias can becircumvented, we draw on the distinction betweentwo cognitive systems—the meaning-based systemand the experience-based system (Allen, Kaut, &Lord, 2008;Epstein,Pacini,Denes-Raj, &Heier, 1996).The meaning-based system processes the meaning ofwords, symbols, and other concepts. It is the part ofthe mind in which people assess data, contemplateideas, and think abstractly. It stores informationaccording to semantic relationships—that is, the ex-tent to which different ideas have similar meanings.People use the meaning-based system to understandthat the word “organization” is closely related to theword “company,” but not to the word “pronounce.”By contrast, the experience-based system processes

6 An image-based vision boosted team creativity whenleaders complemented itwith a focusedvalue system (akinto howMartin Luther King, Jr. centered his images on onecore value—equality) so that people not only visualize thefuture in the same way, but understand the underlyingmeaning of these images in the sameway.However, Cartonet al. (2014) found that imagery in vision communicationhas a direct influence on a shared sense of purpose andcoordination.


sensory information and underlies peoples’ ability toimagine and visualize events. It is the part of themindin which people encode the size of the room they arein, the colors of the clothes they arewearing, andwhois sitting next to them. It is also responsible for auto-biographical memories, such as recollections thata personmight have of the interior of a previous homeor a scenic view from a past vacation. We will nowargue that the blurry vision bias (1) exists becauseleaders overemphasize the meaning-based system,and (2) can be overcome by helping leaders activatethe experience-based system by imagining what theworldwill look, sound, and feel like if andwhen theircompany’s vision is achieved.

How the Blurry Vision Bias Stems from anOveremphasis on the Meaning-based System

The blurry vision bias arises in large part fromhowpeople tend to think about the future. Given that thefuture has not yet transpired, it cannot be seen or felt.Consequently, as individuals contemplate the future,they tend to rely on an abstract understanding ofwhatit willmean (“providing excellent customer service”)instead of a mental simulation of how it could beexperienced (“making customers smile”) (Trope &Liberman, 2003). This tendency is pronounced forthose who are most likely to craft and communicatevisions—upper echelon leaders—becausepeoplewhopossess power tend to think in broad terms about theorganization’s strategy (Magee, Milliken, & Lurie, 2010;Rucker,Galinsky,&Dubois,2012;Smith&Trope,2006).

We argue that the predisposition to think about thefuture abstractly rather than vividly is exacerbatedwhen leaders are confrontedwith theneed tonot onlythinkabout the future, but communicate about it. Thisis because the primary medium of communication—language—tends to reinforce abstract thinking. Thelanguage-processing center is rooted in the meaning-based system because it involves the comprehensionof symbols (letters and words) (Tulving, 1972). Oncethis system is engaged, people seek to “make mean-ing” by building an association between one conceptand a second concept in order to improve their un-derstanding of the first concept (Pratt & Ashforth,2003).Critically, the secondconceptneednotbevividto impart meaning to the first concept (Rosso, Dekas,& Wrzesniewski, 2010). To illustrate, a leader of anonline social networking company can help em-ployees make sense of the future with an abstractphrase such as “helping people make connections.”This phrase lacks imagery because it does not de-scribe anobservable action, but a leader can still use it

to understandwhat the company aims to achieve. It iscertainly possible that a leader could use rhetoric thatpossesses imagery to better understand this com-pany’s future (e.g., “people exchanging the firstwordsof what will be a lifetime friendship”); however, im-agery is not necessary to satisfy the meaning-basedsystem’s need for comprehension. Since people aremore likely to generate and communicate image-based rhetoric when the experience-based system isengaged—yet a focus on language does not activatethe experience-based system—leaders who are notexplicitly nudged to engage the experience-basedsystem are unlikely to employ image-based rhetoricwhenarticulating a vision. The ironic upshot of this isthat leaders who focus on the very medium throughwhichvisionsarecommunicated (wordsandphrases)remain entrenched in the meaning-based system andreinforce their inclination to convey the abstractmeaning of the future, rather than provide an image-laden portrait of it.

Although an excessive focus on language exacer-bates the blurry vision bias because it causes leadersto overweigh the meaning-based system, dominantperspectives in the literature have assumed thatcareful word selection is paramount for effective vi-sion communication. In a review of 180 sourcescontaining advice on how to construct visions,7 wefound that 175 (97.3%) encouraged leaders to care-fully attend to the selectionofwords andphrases (seeAppendix B at blurryvisionbias.wordpress.com formore detail). Although this is not surprising becauselanguage is perhaps the most fundamental charac-teristic of vision communication (Awamleh &Gardner, 1999), it is problematic because it re-inforces the blurry vision bias. More paradoxically,even expert recommendations that explicitly callleaders’ attention to the very category of languagethat elicits imagery (e.g., “use words with lifelikedetail”) are unlikely to counteract the blurry visionbias because they still target the meaning-basedsystem. Consider a leaderwho aims to create a visionfor a government agency and is countenanced by an

7 Since the literature on vision communication is vast, itis challenging to review it systematically. Therefore, weintroduced a new approach to reviewing the literature:semantic cluster analysis. This approach provides away toreview a body of research comprehensively (capturing thefull spectrum of arguments from different articles andbooks), parsimoniously (grouping sources that presentsimilar arguments under the same broad themes), andrepresentatively (establishing the number of sources thatfall into each broad category).



expert to “use words that reflect observable reality,such aswhat you can see and feel.”By being asked tofocus on words that capture observable reality, thispersonwill be prompted to contemplate themeaningof “observable reality” as a concept, which typicallybrings abstract ideas to mind (e.g., “helping thecountry”). Although it is possible that this leaderwould understand “observable reality” via image-based rhetoric (“when people tell their friends theyfeel optimistic about their lives”), such rhetoric is notnecessary to convey this idea’s meaning.

Further compounding this problem are the wordsand phrases experts use to describe image-laden rhet-oricwhen they recommend that leaders use it in visioncommunication. Our literature review (see AppendixB at blurryvisionbias.wordpress.com) indicates thatexperts talk about the general properties that unitedifferent examples of image-laden visions (e.g., “aneffective vision. . .creates a vivid image”). Ironically,although a termsuchas “vivid image”unites countlessimage-based phrases (e.g., a shimmering screen, a ju-bilant customer, a sold-out stadium), the term is itselfa generic, abstract label rather than one that brings tomind an easy-to-visualize person, object, or action(Guadagno et al., 2011; Rosch, 1978). Advice centeredon generic themes will reinforce the tendency forleaders to think about the future at a broad level ofabstraction. A leader who is encouraged to createa “vivid image” is thus more likely to tell listenersabout general properties of the future (e.g., “devicesthat advance mobile technology”) than walk listenersthrough a scene ladenwith visual detail (e.g., “devicesthat complete hundreds of tasks—from ordering foodto starting your car—with the push of a button”). Thenotion that people think of words at the same level ofabstraction aswords they heard justmoments before isconsistent with evidence on the inertia of the brain-storming process, such that new ideas are offshoots ofprevious ones (Berg, 2014). In this way, advice thatfocuses leaders’ attention on general properties ofrhetoric—even image-based rhetoric—will reinforce,rather than counteract, their predisposition to com-municate abstractly about the future.

In sum, leaders are not inclined to craft visionswith imagery, and even leaders who consciouslyseek to infuse their visionswith image-laden rhetoricare unlikely to do so.

Resolving the Blurry Vision Bias by Activating theExperienced-based System

We posit that leaders will formulate image-basedvisions most easily when they engage in a two-step

process: drawing on the experience-based systemandthen having this form of cognition “spill” into themeaning-based system by driving the retrieval ofimage-based rhetoric. First, rather than focusing onselecting words that meet specific criteria, leadersvividly imagine future experiences (Path A inFigure 1). They can do this by mentally projectingthemselves to a day in the future and envisioningwhat the world will look like—such as the shape ofbuildings, the colors of background scenery, and theexpressions on people’s faces. This act of psycholog-ically projecting across time, or temporal projection,is a prospective tactic related to mental time travel(Suddendorf et al., 2009), temporalmental simulation(Waytz, Hershfield, & Tamir, 2015), and episodicforesight (Bulley, Henry, & Suddendorf, 2016), exceptthat people focus less on how they will personallylook and feel in the future and more on what theirorganization will look like, as well as how the worldwill change if their organization’s vision is realized.Further, our proposed intervention is distinct fromthe above-cited tactics because it is used to enhancecommunicationand interpersonal influence, rather thanpeople’s private introspection and planning.

After individuals construct mental imagery aboutthe future and are then asked to describe it to em-ployees, they will create words in the meaning-basedsystem (because that system houses the languagecenter); however, they are likely to find it easiest to useimage-based rhetoric, given that abstract words willnot convey thevisual snapshot theysee in theirmind’seye (PathB inFigure 1). In short, the experience-basedand meaning-based systems “cooperate,” such thatleaders begin the process of constructing a vision bygenerating mental imagery and only then search forthe words to clairvoyantly describe it. Thus, leaderswhopsychologically skip forward to thedistant futurebefore crafting and communicating a vision will openup the boundary between the experience-based andmeaning-based systems of cognition.

To illustrate, consider a leader of a social net-working company who is asked to “imagine peopleusing yet-to-be-created social networking technologyin 10 years.” This guidance pushes the leader topsychologically transport to the future and summona mental image of what it could look like. Once thisimage has been formed and the leader attempts todescribe it, he or she can best do so with image-basedwords (e.g., “a father in the military talking to hischildren back home” or “women from neighboringuniversities typing words that will inspire a socialmovement”). Abstract words would be insufficientfor describing a mental image. Along these lines, we



expect that leaders who deliberately take a mentalleap forward to amoment in the distant future (whichwe hereafter label as both “temporal projection” and“mental time travel”) will employ more image-ladenrhetoric than even those who consciously set out toconstruct visions with such rhetoric.

Hypothesis 1. Leaders who mentally project them-selves to the distant future (i.e., temporal projection)will use more image-laden rhetoric in vision com-munication compared to leaders who emphasizeword selection.

Although this hypothesis relates to the quantity ofimage-based rhetoric, in our studies we account notonly for the amount of image-based rhetoric, but thequality of the images. In addition, we assess overall

vision quality, which, according to our literaturereview, is driven not only by whether a vision pos-sesses imagery, but also whether it is specific,achievable, and reflects the core values and identityof the organization and its members. Although it isoutside the scope of this paper to construct hypoth-eses about the impact of temporal projection on thesethree dimensions, in our studies we assess its effecton all dimensions of vision quality. If leaders whomentally project themselves deep into the future areable to boost one core dimension of vision quality(imagery) while not harming—or even whileboosting—the others, then it will yield a net benefitsuch that they will construct and communicatehigher-quality visions compared to leaders whomake deliberate attempts to boost vision quality

FIGURE 1Theoretical Model

Prescription Type

Language-Centeredversus

Temporal Projection

Cognitive Style

Efficient Thinkersversus

Analytical Thinkers

Cognitive Processing

Meaning-Basedversus

Experience-Based

Vision Quality

AB

Notes: The solid lines represent direct effects: the main effect of prescription type on vision quality (Hypothesis 1) and the interactionbetween prescription type and cognitive style on vision quality (Hypothesis 2). The dashed lines represent the conditional indirect effect(Hypothesis 3) that explains the Hypothesis 2 interaction. In line with Hayes (2013), the interaction on the second stage should only beinterpreted in tandem with the path between temporal projection and the mediator.


through careful word selection (i.e., tactics that en-gage the meaning-based system). As another way ofapproximating overall vision quality, we assesswhether the visions leaders craft as a result ofmentaltime travel are effective enough to increase the extentto which their employees are inspired at work.

The Moderating Role of Cognitive Style

The link between temporal projection, mental im-agery, and imagery in vision communication is likelyto be influenced by an individual difference pertain-ing to the willingness to accommodate experience-basedprocessing (Cacioppo, Petty, Feinstein,& Jarvis,1996; Epstein et al., 1996). Although this individualdifference exists as a continuous dimension, it is bestappreciated by contrasting its two poles: whereas ef-ficient thinkers prefer not to engage in over-analysis,analytical thinkers are methodical and have a strong“need for cognition.” Efficient thinkers are likely tolet mental imagery freely influence the way theycommunicate. They are not inclined to overrideimaginative thinking with controlled logic, supplantsnapshots that they envisage in their mind’s eye withsemantic analysis, or impute anabundanceof abstractmeaning into their mental glimpses of the future(Epstein et al., 1996). That is, efficient thinkers arecomfortable focusing on a single graphically definedscenario, rather than forming general impressions. Assuch, we expect that efficient thinkers will let vividimages loom large in their consciousness and guidetheir search for words and phrases when they seek tocommunicate their thoughts to others.

In contrast, analytical thinkers are more inclinedto consciously override the influence of mental im-agery on how they communicate. Since they prefercomplex problem solving and effortful thought, theytend to think at a high level of abstraction. By defi-nition, abstract thinking involves drawing connec-tions across disparate domains and contemplatinggeneral principles, rather than specific events (Rosch,1978). Therefore, even if they are capable of vividlyimagining a single future event, analytical thinkersare likely to explain their understanding of the futureusing abstract categories that represent the broadermeaning of that event (Kahneman, 2011). For in-stance, an analytical thinkerwho imagines customerssmilingmayconsider thebroadermeaningof this typeof scenario and craft a vision in which “customers arealways satisfied.” In this way, a natural dispositiontoward abstract cognitive processing leads analyticalthinkers to favor concepts, rather than words thatpertain to sensory experiences.

In short, since they prefer to “abstract out” tomakegeneral summations, analytical thinkers are likely topossess a stronger barrier between the experience-based and meaning-based systems compared toefficient thinkers. Thus, exercises that triggerexperience-based thinking (e.g., mentally projectingoneself into the distant future) are less likely to“spill” into the meaning-based system for analyticalthinkers than for efficient thinkers. In this way, thegeneration of mental imagery via a vivid mentalsimulation of thedistant future is less likely to lead tothe communication of imagery via image-basedrhetoric. This is a critical distinction becauseleaders can only influence others through explicitcommunication. In short, analytical thinkers are notlikely to be as amenable as efficient thinkers will beto letting temporal projection influence theircommunication.

Hypothesis 2. Individual cognitive style (efficientversus abstract thinking) and prescription type (tem-poral projection versus language-based prescrip-tions) will interact, such that temporal projectionwill boost the use of image-laden rhetoric in visioncommunication more for efficient thinkers than foranalytical thinkers.

Altogether, our logic leads us to predict that therelationship between temporal projection andimage-based rhetoric will be mediated by theexperience-based system and moderated by cogni-tive style, such that the relationship between tem-poral projection, mental imagery, and imagery invision communication will be stronger for efficientthinkers than for analytical thinkers (see the dashedlines in Figure 1). That is, analytical thinkers over-ride the link between the generation of mental im-ages and the communication of images to a greaterextent than do efficient thinkers.

Hypothesis 3. The role of mental imagery as a medi-ator between prescription type (temporal projectionversus language-based) and image-laden rhetoricwillbe moderated by cognitive style, such that the medi-ated relationship will hold more for efficient than foranalytical thinkers.

METHOD

We tested the hypotheses in three studies. Weaimed to promote external validity by recruitinga diverse set of participants, including a samplewithBritish government officials and a sample with se-nior corporate executives. All three studies wereexperiments, thereby establishing internal validity.


In study 3 we assessed whether the visions that se-nior executives craft as a result ofmentallyprojectingthemselves to amoment in the future are superior forinspiring their own employees. Thus, we establishpredictive validity not only for our primary inter-vention (temporal projection / mental time travel),but also for imagery in vision communication. Fi-nally, in addition to testing our predictions threetimes, we sought to establish the replicability of ourfindingsbypreregistering thehypothesesandmethodsfor studies 2–3 and running one high-powered designwith 700 full-time employees (study 2). All data areavailable from the first author.

Study 1: Brexit Experiment

Sample and design. We sought a context thatwould establish ecological validity, such that visioncommunication was timely and meaningful. Imag-ery in vision communication can be especially use-ful during periods of change because it crystallizesthe ultimate goal of a change effort (Heath & Heath,2010; Kotter, 1996). For this reason, we launchedan experiment involving British government em-ployees on June 24, 2016, the day the United King-dom announced that it would exit the EuropeanUnion (“Brexit”) as the result of the vote of Britishcitizens.8 The Brexit vote marked the beginning ofa transitionmany experts considered more dauntingthan any theU.K. government had experienced sincethe end of World War II. The day of the vote wasknownwell in advance, allowing us to plan the exacttiming of the launch of the experiment. In fact, webegan the experiment just a few hours after DavidCameron announced his resignation as prime min-ister (the highest ranking office in the Britishgovernment).

The sample consisted of 166 government officialsfrom the United Kingdom. The results remained sub-stantively the same when retaining nine participantswho failed an attention check, so we included them toincrease statistical power. The sample included coun-cil members from provincial governments, heads ofdepartment, executive officers, education officers,national parks managers, heads of communicationsystems, administrators in the National Health Ser-vice, and principle engineers. We identified partic-ipants via an online database (prolific.ac). Theaverage participant age was 37, with an average of11.3 years of experience, and 66% were female. A

number of participants commented after the experi-ment was over about the relevance of Brexit to theirjobs. We were explicit that the timing of the studywas due to Brexit and we emphasized the govern-ment transition that would take place. Beyond theopportunity to examine a context involving change,we sampled government employees because visioncommunication is viewed as essential among gov-ernment leaders, who must unify various stake-holders. Despite the unique context, our aim was toensure that participants focused on crafting a visionfor an organizational purpose in addition to a politi-cal purpose. Thus, we emphasized that participantsshould craft a vision relevant to both the Britishgovernment overall as well as their own company,agency, or unit.

Procedure. We constructed six conditions ina between-subjects design. We did not drop anycases from any of these conditions or perform ana-lyses before the data were collected, and we stoppeddata collection only when no further participantswhomet our sample requirements volunteered. Onecondition was a control group, in which we askedparticipants to write a vision statement without anyprompts or prescriptions. The next four conditionsfeatured what we call language-based prescriptions,because they involved explicitly instructing leadersto incorporate certain categories of words into theirvisions of the future. As noted earlier, scholars haveconverged on four desirable attributes in a vision:imagery, specificity, achievability, and values. Par-ticipants were explicitly asked to choose words thatreflected these attributes. We ensured that thewording of the prescription in each condition in-cluded the most frequently invoked words thatauthors across the 180 sources identified in our lit-erature review used to describe each attribute. Thisensured that each prescription was confined to aunique, delineated domain. We constructed eachprescription to be similar in word count. For theimagery / vividness prescription we noted that “apicture is vivid when it is depicted as if it could beseen and felt. It should simulate a visual represen-tation or an image of an event. A vivid statement ispicture-like. It portrays action, events, and objectswith lifelike detail.” For the specificity prescription,we explained, “Specificity refers to a statement’sclarity and exactness. A specific statement is dis-tinct, easy to understand, and can be interpreted inonly one way. Statements that are simple are oftenclearer. A vision that is specific is free from ambi-guity.” For the achievability prescription, we stated,“Achievability refers to whether a statement

8 British citizens voted on Brexit the day before the de-parture was announced (June 23, 2016).


describes an end-state or event that can actually beattained. An achievable vision is a vision that canone day be realized. That is, it can actually be ac-complished in the future.” For the values pre-scription, we noted, “Your statement should reflectyour values (standards for what is desirable). Itshould communicate a core ideology—what is bothmeaningful and important. It should convey an ide-alized goal that directs action. Those who hear yourvision should be motivated and inspired. Theyshould aspire and dream to attain the vision.”

The final condition involved an experimentalmanipulation of our core construct, temporal pro-jection. We asked participants to psychologicallyproject themselves deep into the future and con-struct a mental picture of what a moment in thedistant future could look like. We adapted theexperience-based prescriptions found in our break-down of the existing literature (see Appendix B atblurryvisionbias.wordpress.com). As with the otherconditions, we accounted for the government’smultiple layers by first asking participants to makesure they created a vision that was aligned with thegoals of the entireBritish government.We then askedthem to customize the vision to the specific entity inwhich theyworked so that it would bemotivating forthose in their immediate working environment.Specifically, they were told, “Imagine you entera time machine and emerge in the future just afteryour company has achieved its vision. What doesthis future look like? Take a picture with your cam-era. Think of how to make your vision embody whatyou saw in the picture you took.”9 Although timemachines do not exist, the timemachine interventioncombines a hypothetically credible and psychologi-cally real experience that utilizes the cognitive pro-cesses needed to envision the future (Schacter,Addis,& Buckner, 2007). We label this intervention the“mental time travel condition,” or the “timemachinecondition.”

Measures. Consistent with our theoretical model,the mediator (mental imagery) was measured afterthe manipulations but prior to when participantscrafted visions. We adapted items from research onmental imagery (Babin & Burns, 1998) (e.g., “Rightnow there is a visual scene playing in my ‘mind’seye’;” 1 5 disagree to 7 5 agree; a 5 .90). To assessour moderator (analytical versus efficient thinking),we employed the scale created by Epstein et al.(1996) in their validation of an analytical thinkingmeasure. An example item is, “I prefer complex tosimple problems” (15disagree to 55 agree;a5 .66).As noted in our theorizing, this construct is unidi-mensional, such that people range from highly effi-cient to highly analytical in their thought style.

For the primary dependent variable, image-basedrhetoric,we employed theCoh-Metrix dictionary, anonline database of 4,825 words coded by adult ratersaccording to the extent to which each word repre-sents an image (Graesser, McNamara, Louwerse, &Cai, 2004). Each word was assessed from 100 (weakimagery) to 700 (strong imagery). The imagery scorewas the average amount of imagery per word in eachvision. This measure has several advantages. It hasbeen psychometrically validated. Further, it opti-mizes reliability—since it is automated it codes eachappearance of the same word in the same way. Re-liability is further enhanced given that each word inthe dictionary has been coded by thousands of adults.This measure is especially appropriate given that in-dividual words are the “raw material” for imagery.That is, a statement cannot elicit mental imagery if itcontains abstract words (e.g., “protocol”) rather thanimage-laden words (e.g., “smile”). In a subsequentsection (entitled “Supplementary Analyses”) we re-port the results of analyses with an alternate measureof this dependent variable to ensure that our resultsare robust. This alternatemeasure uses human codersto assess the overall imagery of the higher-ordercomponents of each phrase (clauses and the entirestatement) and allowed us to retain the validity andreliability of the automated codingweare using inourprimary analyses while increasing confidence thatthe results would translate to individuals hearing anentire statement from a leader.

We then assessed the three other features scholarshave identifiedascentral tovisionquality (specificity,achievability, and values). We used the Coh-Metrixdimension hypernymy to measure specificity (seeCrossley,Salsbury,&McNamara [2009] forhowthis isaccomplishedvia categorizationhierarchies). Visionswere achievable to the extent that they could beattained. Given that this could only be inferred at the

9 To determine whether the word “company” was ap-propriate to represent the government entities in oursample, we conducted three tests. Prior to running theexperiment, two British citizens reviewed all wording andagreed that “company” was an appropriate term. We thenfound that 100% of a sample of 81 government employeesfromBritain noted that theword “company” is appropriatefor describing a government entity in response to a “yes orno” question. Third, we conducted an archival analysis,which suggested that this word is used to capture variousentities (e.g., “fire company,” “military company”), espe-cially in Britain. This word also generalizes well to thesamples we use in studies 2 and 3.



statement level, two coders blind to the study hy-potheses coded this dimension on a scale from 100 to700 (a 5 .65). Because a small number of values isideal (Carton et al., 2014), for each statement wecounted the number of times participants invoked thenine values that practitioners most frequently use(Carton et al., 2014;Hansen, 2010;O’Reilly, Chatman,& Caldwell, 1991). We checked to ensure that theimagery participants created was congruent with thevalues they invoked.

Controls. In a questionnaire, we accounted for thepossibility that the mental time travel condition hasmore powerful effects on vision quality because it isamore intrinsicallymotivating task, using a four-itemscale from Grant (2008). We also accounted for par-ticipant engagement by measuring the time partici-pants spent on the task during the unbroken periodafter participants were all exposed to the manipula-tion. To test for the possibility that the mental timetravel condition yielded positive effects because theinstructionswere easier to follow,we included a four-itemmeasure of cognitive fluency, adapted frompriorfluency research (Lucas & Nordgren, 2015; Schwarz,Bless,Strack,Klumpp,Rittenauer-Schatka,&Simons,1991). We also accounted for an individual’s level inthe organizational hierarchy. Finally, we assessedeach participant’s power via the number of direct re-ports he or she oversaw.10

Results and discussion. Correlations and de-scriptive statistics are provided in Table 1. We in-cluded the control variables in all models; however,the results remained substantively the same whencontrols were not included. We analyzed the data us-ing analysis of covariance and ordinary least squares(OLS) regression (see Table 2 for regression results).

In support of Hypothesis 1, the mental time travelcondition (M 5 406.902, SD 5 28.398) boostedimage-laden rhetoric in vision communicationmorethan did any other condition (M 5 395.383, SD 524.259), F(1,159) 5 5.663, p 5 .019, h2

p 5 .034. TheR2 for the model was .062. To further parse theseresults, we constructed a coding scheme with sepa-rate fixed effects for: (1) all four language-basedconditions, and (2) the no-instruction control con-dition with the mental time travel condition as thereference category (Cable, Gino, & Staats, 2013), andfound that mental time travel significantly increasedimagery in vision communication compared toall four language-based conditions, F(1,158) 5 5.19,

p 5 .024, h2p 5 .032, as well as the no-instruction

control condition,F(1,158)5 3.91,p5 .049,h2p5 .024

(see Figure 2a). Mental time travel also induceda greater amount of average imagery compared toeach individual condition, including the conditionin which leaders were explicitly instructed to usevivid language, and a model including fixed effectsfor each condition established that the mental timetravel condition was the only one of the five experi-mental interventions that induced significantlymore imagery compared to the control condition.Wealso tested the effect of the mental time travel con-dition on ourmeasures of the three other elements ofvision quality (specificity, achievability, andvalues) and did not find differences for any of thesevariables (p. .1 for all variables). Consequently, wecan conclude that temporal projection via mentaltime travel boosts a key aspect of vision quality(image-laden rhetoric) without affecting the otherthree elements, thereby resulting in a net improve-ment in vision quality.

In support of Hypothesis 2, there was a significantinteraction between prescription type and cognitivestyle, F(1,157) 5 12.45, p 5 .0005. The simple slopewas significant for efficient thinkers (b5 16.72, SE55.24,p5 .001), butnot for analytical thinkers (p5 .42).We plotted the data at the 25th and 75th percentiles ofobserved values of themoderator. As shown in the topgraph inFigure2b, efficient thinkers generatedvisionswith significantly more image-laden rhetoric in themental time travel condition than in the language-based conditions, whereas analytical thinkers did notbenefit from mental time travel relative to the otherconditions.

We then tested our prediction thatmental imagerywouldmediate the relationship betweenmental timetravel and imagery in vision communication moreamong efficient compared to analytical thinkers—asecond-stage conditional indirect effect (as shown inFigure 1, we expected that all leaders who mentallyprojected themselves to a moment in the distant fu-ture would experience mental imagery, but leaderswho are analytical thinkers would be more inclinedto override the effect of mental imagery on visioncommunication). We tested the conditional indirecteffect with bootstrapped confidence intervals usingHayes’ (2013) PROCESS. The mediator (mental im-agery) was censored such that a disproportionatenumber of cases were located at the lowest value(about twice as many cases had a value of 1 than anyother value; the distribution was otherwise normal).We accounted for this with truncated regression(Breen, 1996). Bootstrapped confidence intervals at

10 The results reported below remained at the samelevels of significance whenwe included other measures ofhierarchy, including reward power and status.


TABLE1

Studies1an

d2:

Descriptive

Statisticsan

dCorrelation

s

Mea

nSD

12

34

56

78

Study1

1.Visionim

agery

397.46

425

.357

2.Tem

poral

projection

0.18

00.38

5.175

*3.

Men

talimag

ery

3.38

91.68

8.134

*.369

**4.

Analyt.v

s.eff.thinking

2.25

10.66

10.01

2.159

*20.04

25.

Intrinsicmotivation

4.57

81.49

00.06

30.06

5.258

**–.230

**6.

Tim

eon

task

(sec

onds)

112.56

616

2.64

120.03

4.164

*20.00

820.01

20.03

27.

Fluen

cy4.60

11.21

50.07

020.00

4.271

**–.215

**.316

**20.07

48.

Lev

elin

orga

nization

3.72

91.21

8–.158

*0.00

20.05

50.06

720.03

020.01

420.01

99.

Direc

trep

orts

2.93

411

.202

0.06

520.06

220.03

220.04

90.10

720.03

20.03

2–.305

**Study2

1.Visionim

agery

394.21

028

.507

2.Tem

poral

projection

0.15

00.35

5.220

**3.

Men

talimag

ery

4.03

71.57

1.133

**.278

**4.

Analyt.v

s.eff.thinking

2.25

30.85

50.01

5.094

*20.02

15.

Intrinsicmotivation

4.88

61.57

00.01

20.05

4.379

**–.266

**6.

Tim

eon

task

(sec

onds)

123.79

511

1.98

60.00

1–.150

**.070

20.05

0.158

**7.

Fluen

cy5.07

01.11

020.02

920.00

5.075

*–.301

**.303

**0.06

88.

Lev

elin

orga

nization

3.33

00.89

40.02

10.00

720.07

20.05

5–.139

**20.02

50.00

09.

Direc

trep

orts

4.89

012

.358

20.00

20.04

5.112

**20.03

8.081

*0.00

00.02

5–.404

**

Notes:InStudy1,

n5

166an

din

Study2,

n5

652.

Cod

ing:

15

Tem

poral

Projection,0

5other

conditions.

*p,

.05

**p,

.01

***p

,.001


TABLE2

Studies1an

d2:

MainEffec

ts,Interactions,an

dCon

ditional

Indirec

tEffec

ts(H

ypotheses

1–3)

Variable

b(SE)

Variable

b(SE)

Variable

b(SE)

Variable

b(SE)

STUDY1

Study1MainEffec

t(Hyp

othesis

1)Study1Interaction(H

ypotheses

2)Study1Con

ditional

Indirec

tEffec

t(Hyp

othesis

3)

DV:Imag

e-Based

Rhetoric

DV:Imag

e-Based

Rhetoric

DV:M

entalImag

ery

DV:Imag

e-Based

Rhetoric

Intrinsicmotivation

.491

(1.388

)Intrinsicmotivation

–.376

(1.392


.151

(.08

2)Intrinsicmotivation

.664

(1.603

)Tim

eon

task

–.010

(.01

2)Tim

eon

task

–.011

(.01

2)Tim

eon

task

–.001

(.00

1)Tim

eon

task

–.007

(.01

3)Fluen

cy1.11

3(1.693

)Fluen

cy.909

(1.661

)Fluen

cy.218

*(.10

5)Fluen

cy1.87

4(2.072

)Organ

izational

rank

23.12

5(1.677

)Organ

izational

rank

22.84

6(1.629

)Organ

izational

rank

.100

(.10

0)Organ

izational

rank

23.44

5(1.926

)Direc

trep

orts

.053

(.18

4)Direc

trep

orts

.063

(.17

8)Direc

trep

orts

–.003

(.01

0)Directrep

orts

.006

(.18

6)Tem

poral

projection

12.200

*(5.127

)Tem

poral

projection

255

.324

**(19.75

6)Tem

poral

projection

1.30

7***

(.28

4)Tem

poral

projection

7.48

0(6.057

)Analyt.v

s.effic.

24.82

2(3.366

3)Men

talimag

ery

3.92

5(4.712

)Tem

poral

Projection3

Analyt.v

s.effic.

27.948

***(7.922

)Analyt.v

s.effic.

2.29

7(8.631

)Men

talimag

ery3

analyt.v

s.effic.

.077

(2.071

)Indirec

tEffec

t(b)

Indirec

tEffec

t95%

CI

Effic.thinkers

5.60

(1.60,

12.88)

Analyt.thinkers

5.36

(–.17,

12.35)

STUDY2

Study2MainEffec

t(Hyp

othesis

1)Study2Interaction(H

ypotheses

2)Study2Con

ditional

Indirec

tEffec

t(Hyp

othesis

3)

DV:Imag

e-ba

sedRhetoric

DV:Imag

e-ba

sedRhetoric

DV:M

entalImag

ery

DV:Imag

e-ba

sedRhetoric

Intrinsicmotivation

.124

(7.373


.034

(.76


.370

***(.04


21.46

5(.94

2)Tim

eon

task

.009

(.01

0)Tim

eon

task

.010

(.01

0)Tim

eon

task

.001

(.00

1)Tim

eon

task

.010

(.010

)Fluen

cy–.824

(1.035

)Fluen

cy–.877

(1.064

)Fluen

cy.008

(.05

8)Fluen

cy–.647

(1.165

)Organ

izational

rank

.636

(1.349

)Organ

izational

rank

.687

(1.346

)Organ

izational

rank

.082

(.07

5)Organ

izational

rank

.398

(1.457

)Directrep

orts

–.009

(.09

7)Direc

trep

orts

–.015

(.09

7)Direc

trep

orts

.006

(.00

5)Directrep

orts

–.023

(.09

8)Tem

poral

projection

18.063

***(3.134

)Tem

poral

projection

–.767

(9.061

)Tem

poral

projection

.944

***(.16

3)Tem

poral

projection

16.515

***(3.271

)Analyt.v

s.effic.

21.75

4(1.479

)Men

talimag

ery

.437

(2.003

)Tem

poral

Projection3

analyt.v

s.effic.

7.88

6*(3.533

)Analyt.v

s.effic.

25.02

7(4.311

)Men

talimag

ery3

analyt.v

s.effic.

.723

(.865

)Indirec

tEffec

t(b)

Indirec

tEffec

t95%

CI

Effic.thinkers

1.83

(.16

,4.16)

Analyt.thinkers

1.14

(–.35,

2.88

)

Notes:D

V5

Dep

enden

tvariable.In

themaintext

wereporta

nF-testfrom

anANCOVAforHyp

otheses

1an

d2ba

sedon

conve

ntion

interm

sof

how

hyp

otheses

arereportedin

experim

ents;h

owev

er,since

itis

nec

essary

toreportregression

coefficien

tsforHyp

othesis

3,in

this

tablewemaintain

consisten

cyby

reportingregression

coefficien

tsforall

hyp

othesis

tests.

*p,

.05

**p,

.01

***p,

.001


FIGURE 2Studies 1–3: Main Effects and Interactions (Hypotheses 1 and 2)

Figure 2a: Main Effect of Prescription Type on use of Image-based Rhetoric in Vision Communication forStudies 123

370

380

390

400

410

420

Control Language-BasedPrescriptions

Temporal Projection(Mental Time Travel)

Study 1

Study 2

Study 3

FIGURE 2bFigure 2b: Interaction between Prescription Type and Cognitive Style on use of Image-based Rhetoric in Vision

Communication for Studies 1 and 2

Study 1

370

380

390

400

410

420

Language-BasedPrescriptions


Study 2

Language-BasedPrescriptions


370

380

390

400

410

420

Efficient Thinkers

Analytical Thinkers

Efficient Thinkers

Analytical Thinkers

Notes: For all graphs the error bars are 95% CIs. In study 3 the only language-based prescription that was tested involved explicitinstructions to include image-based rhetoric.


the 25th and 75th percentiles of the moderator dem-onstrated that the indirect effect ofmental time travelon leaders’ use of imagery in vision communicationvia the activation of mental imagery excluded zerofor efficient thinkers (indirect effect5 5.60; 95% CI:1.60, 12.88) but not analytical thinkers (indirect ef-fect 5 5.36; 95% CI: 2.17, 12.35). As a robustnesscheck, we corrected for censoring of the mediator asa dependent variable via tobit regression in STATA(Breen, 1996). The results remained substantively thesame, providing further support for this hypothesis.

The Brexit experiment provided support for thehypotheses. This context has advantages in that itis a controlled experiment featuring a situation (ex-treme change) that would serve as an incentive forparticipants to improve their ability to craft andcommunicate visions of the future. However, theunique nature of a sample of British government of-ficials raises questions about the generalizabilityof temporal projection to other contexts. Further, inaddition to assessing vision quality by breakingdown leaders’ visions according to the four quality-based attributes, it would be helpful to assess visionquality by measuring the extent to which the re-cipients of visions perceive them to be inspiring.Additionally, it would be beneficial to assess the ef-fectiveness of temporal projection with a largersample size. To gain greater confidence in our re-sults, we conducted two more studies to establishthese forms of validity and reliability. In Study 2 wetested our hypotheses in a high-powered replicationstudy, aswell aswith five subsamples of leaderswithvarying numbers of direct reports. In Study 3 wetested temporal projection with a sample of corpo-rate executives and assessed the extent to which theexecutives’ employees were inspired by the visionsthey crafted as a result of psychologically leapingforward to a moment in the future.

Study 2: Replication with High-powered Designand Leaders with Large Spans of Control

We first sought to increase confidence in the rep-licability of our results by conducting a follow-upstudy that included key adjustments in line withrecommendations by Button et al. (2013) andMunafoet al. (2017). Before collecting data, we preregisteredour sample parameters, sample size, hypotheses,conditions, and measures at aspredicted.org (pleasesee https://aspredicted.org/5av33.pdf for the officialdocument). Further, we recruited a larger sample(100 full-time employees per condition) to increaseconfidence in the replicability of the findings

(Simmons, Nelson, & Simonsohn, 2011). This sam-ple size also enabled us to isolate subsamples ofleaders with different numbers of direct reports,which could be illuminating since leaders are likelyto wieldmore influence and have greater experiencecrafting visions as they oversee a greater number ofemployees.

Design, sample, and procedure. We recruited700 full-time employees across seven conditions. Inaddition to the six conditions featured in Study 1,weadded a seventh condition: leaders who are given allfour language-based (i.e., meaning-based) pre-scriptions at the same time.11 We accessed full-timeemployees through a filter on Mechanical Turk,a population shown to provide data quality superiorto university labs and equivalent to many work set-tings (Buhrmester,Kwang,&Gosling, 2011; Paolacci,Chandler, & Ipeirotis, 2010). Part-time employeesand unemployed workers could not participate.Participants came from various countries (includingChina, Honduras, India, the Netherlands, Philip-pines, Singapore, Ukraine, and the United States).The average age was 35, and 48% were women. Toboost generalizability, participants created a visionfor one of two industries—either financial services orhealth services. Forty-four participants who did not

11 In the first two studies we compared temporal pro-jection to four separate conditions, such that each condi-tion reflected one of the language-based (meaning-based)prescriptions.Wedid this in order to (1) reflect the existingliterature, in which we found that over 80% of expertsrecommended only one of the four approaches, therebyboosting our method’s ecological validity, and (2) ensurethat the amount of advice we gave to participants in eachcondition was the same. In terms of this latter issue, thewordcount for eachconditionwas constrainedbyourneedto include words and phrases that have been used in theliterature (see Appendix B at blurryvisionbias.wordpress.com) so that our instructions represented how scholarsdescribed each of the four themes. The mental time travelcondition (47 words) was thus similar in length to each ofthe other four treatment conditions (which averaged 43words), and therefore the prescriptions differed in terms ofcontent, and not volume or detail. In contrast, all fourlanguage-based prescriptions put together would be 172words long—four times as long as the mental time travelprescription. However, including such a condition wouldrule out the possibility that leaders who are given moreadvice (all four language-based prescriptions at once) willgenerate better visions. Finally, wemademinor changes tothe wording of the mental time travel condition in order tofurther improve its fidelity to the literature and to thetheoretical construct; see Appendix B (blurryvisionbias.wordpress.com) for the exact wording.


http://aspredicted.org

https://aspredicted.org/5av33.pdf





complete the task, and four that failed an attentioncheck, were excluded from the analyses, as specifiedin the preregistered document, leaving a final sam-ple of 652. However, the results we report belowremained substantively the same when these par-ticipantswere included.Weused the samemeasuresas in the Brexit experiment. The scales were reliable:a5 .94 formental imagery (themediator) anda5 .83for cognitive style (the moderator). We once againused ratings of achievability by two blind un-dergraduate coders, following the same procedure(a 5 .65).

Results. Using analysis of covariance (ANCOVA),we again found support for our first hypothesis.Participants in themental time travel condition (M5409.269, SE 5 33.25) communicated visions withsignificantly more image-laden rhetoric than thosein the other conditions, (M 5 391.611, SD 5 26.80),F(1,645) 5 33.219, p 5 .00000001, h2

p 5 .049. Fur-ther analyses revealed that the mental time travelcondition yielded visions with significantly greaterimage-laden rhetoric than did the control andlanguage-based conditions when tested separately(see Figure 2a).12 The mental time travel conditionalso induced significantly more imagery comparedto each individual condition, including the condi-tion in which leaders were explicitly asked to usevivid language, and a model including fixed effectsfor each condition established that the mental timetravel condition was the only one of the six experi-mental interventions that induced significantly moreimagery than the control condition. Regarding effectsonother aspects of visionquality, leaderswhowere inthe mental time travel condition also crafted visionsthat were more achievable (p 5 .0000000000001).We did not find differences between conditionsfor specificity or values. The R2 for the modelwas .051.

Consistent with Hypothesis 2, the interaction be-tween temporal projection and cognitive style (ana-lytical versus efficient thinking) was significant,F(1,643)5 4.981,p5 .026.As depicted in the bottomgraph in Figure 2b, the amount that efficient thinkersincreased their use of image-laden rhetoric in thetime machine condition versus the conditions withlanguage-based prescriptions (b 5 23.13, SE 5 3.94,p 5 .00000001) was greater than that of analyticalthinkers (b 5 9.83, SE 5 4.71, p 5 .037). We thentested for moderated mediation, such that mentalimagery would mediate the relationship betweenmental time travel and image-laden rhetoric moreamong efficient thinkers than among analyticalthinkers. Similar to the Brexit experiment, the me-diator (mental imagery) was censored such thata disproportionate number of caseswere representedat the lowest value (the value of 1 was representedalmost twice as frequently as the next highest value,5, on the scale from 1 to 7; the distribution was oth-erwise normal). We again accounted for this withtruncated regression (Breen, 1996). Providing sup-port for this hypothesis, 95% bootstrapped confi-dence intervals at the 25th and 75th percentiles of themoderator demonstrated that the indirect effect ofmental time travel on image-laden rhetoric throughthe activation of mental imagery excluded zero forefficient thinkers (indirect effect5 1.83; 95%CI: .16,4.16) but not analytical thinkers (indirect effect 51.14; 95% CI: 2.35, 2.88). As a robustness checkwe corrected for censoring of the mediator as a de-pendent variable via tobit regression in STATA(Breen, 1996). The results remained substan-tively the same, providing further support for thishypothesis.

We then examined whether this effect held forleaders with different spans of control (operational-ized as the number of direct reports). Span of controldid not significantly moderate the effect of temporalprojection on image-based rhetoric (p 5 .90), sug-gesting that temporal projection is robust to a leader’sspan of control. To illustrate, temporal projectionsignificantly increased imagery in vision communi-cation relative to the other conditions for leaderswith the following number of direct reports: 10 ormore (n5 95,p5 .004), 20 ormore (n5 41,p5 .048),30 ormore (n5 22, p5 .005), 40 ormore (n5 15, p5.035), and 50 or more (n 5 11, p 5 .045).13

To gather preliminary evidence related towhetherimagery in vision communication has useful

12 To further parse these results, we constructed an al-ternate coding schemewith separate fixed effects for (1) theno-instruction control condition, (2) all four language-based conditions independently, and (3) all four language-based prescriptions at the same time, with the timemachine condition as the reference category (Cable et al.,2013). The time machine condition caused leaders tocommunicate visions with significantly more image-ladenrhetoric compared to those in the no instruction controlcondition, F(1,643) 5 22.775, p 5 .000002, h2

p 5 .034,those who received the four language-based conditionsindependently, F(1,643) 5 28.999, p 5 .0000001, h2

p 5.043, and those who received the four language-basedprescriptions at the same time, F(1,643) 5 20.628, p 5.000007, h2

p 5 .031.

13 Given the sample sizes we excluded covariates in or-der to preserve an acceptable variable-to-observation ratio.


downstream consequences, a health care employeerated the inspirational power of visions created byleaders of the 25 largest units among participantswho crafted visions for health care services. We didnot discuss the coding with the rater after the ini-tial training, review the codes with her, or ask herto recode any of the statements. Rather, we analyzedonly the results she sent us the first time. Inspiration,which was rated on a scale from 100 to 700, waspredicted by leaders’ use of imagery in vision com-munication (r 5 .513, p 5 .012). We followed up onthese preliminary results in Study 3 by assessing theeffect of leaders’ visions on their employees, therebyestablishing that the time machine condition notonly increases leaders’ use of image-laden rhetoricin vision communication, but that this use of im-agery enhances the inspiration of the leaders’ ownemployees.

Study 3: Replication with SeniorCorporate Executives

We sampled 123 senior corporate executives whowork at major firms. In addition to establishing in-ternal validity by again using an experimental for-mat, this served as another ecologically valid contextbecause senior executives have broad influence.Participants crafted vision statements for their owncompanies, thereby increasing external validity.Since we anticipated that it would be challenging torecruit a large number of upper echelon leaders,we compared the two most important experimentalconditions between subjects (the time machinecondition versus the condition inwhich participantswere explicitly instructed to employ visions withimagery). Additionally, prior to the manipulationparticipants constructed a vision without any in-structions, allowing us to use it as a within-subjectscontrol and to account for random effects for eachsubject. Thiswas helpful because unique participant-level variationwas introduced in terms of the contentof the visions since all leaders crafted visions for theirown companies (rather than for the same contentdomains, as in the prior two studies). We again pre-registered the details of the experiment prior to datacollection (please see https://aspredicted.org/39tj7.pdf for the official document).

We recruited participants over a five-month pe-riod.We required that individuals occupy one of thetop few positions in either their company or a largedivision. The sample included CEOs, COOs, CTOs,vice presidents, heads of department, partners, re-gional managers, and firm owners (average age: 37,

42% female). The leaders’ average company sizewas10,610 employees, and the averagenumber of peopleeach leader directly oversaw was 28, although theyoften indirectly oversaw a much larger set of em-ployees because hundreds or thousands of em-ployees were formally under their rank.We plannedto include the full set of questionnaire items, in-cluding those related to themediator andmoderator,but we initially got a limited response rate (pre-sumably because of upper echelon executives’ timeconstraints). To ensure an adequate response ratewhen working with a sample of executives, we fol-lowed Ashford, Wellman, Sully de Luque, DeStobbeleir, and Wollan (2017) by paring down sur-vey length. Thus, the only part of the experiment weadministered for all cases was the experimentalmanipulation—a limitation mitigated by our testingof the mediator and moderator in both studies 1 and2. (In the discussion section we explain that this isnot likely a practical limitation because leaders whowant to employ thismethodwill only need to engagein temporal projection via the timemachine exerciseand not fill out a questionnaire.) We also gave theoption to leaders to send their visions to a coworkerto assess the visions’ inspirational power (n 5 100;average age: 38, 43% female). We used the scale oninspiration from Thrash and Elliot (2003)14 and thesame measures of imagery and vision quality asstudies 1 and 2. Interrater reliability for ratings ofachievability was acceptable (a 5 .70).

Given that each executive crafted two visions (acontrol vision and a treatment vision), we analyzedthe data using linear mixed models, which accountsfor random effects of each participant—an approachconsidered superior to other methods for assessingrepeated measures because it makes fewer assump-tions and leads to less biased estimates, includingthose of the variance-covariance matrices (Bell &Jones, 2015). Further, it permits a direct test of the

14 In 65 cases we used the full four-item scale (a5 .928).In the remaining cases we could only use a single itembecause of participants’ time constraints, and thusweusedthe first of the four items since it was the most represen-tative. Among those who responded to all four items, thisitem had a .900 correlation (p , .00000001) with the fouritem scale, enabling us to analyze all participants together(Ashford et al., 2017). We ran analyses in two ways to en-sure that our results were robust—on the subsample whoresponded using the full four-item scale and with the fullsample controlling for cases that had one item versus thosethat included all four. The results were robust to both ap-proaches (thep values for all effectswere less than or equalto .001 in all cases).


https://aspredicted.org/39tj7.pdf

https://aspredicted.org/39tj7.pdf

mediating role of image-laden rhetoric in the re-lationship between temporal projection and em-ployee inspiration. Consistent with our predictions,when including fixed effects for the control condi-tion and the language-based condition with themental time travel condition as the reference cate-gory (Cable et al., 2013), we found that executiveswho employed mental time travel communicatedvisions with more imagery compared to visionscreated by executives who were explicitly told toinclude imagery, b 5 219.34, t 5 23.66, p 5 .0003andmore imagery compared to visions created in thewithin-subjects control condition, b 5 221.08, t 524.65, p 5 .00001 (see Figure 2a). The effect ofmental time travel versus the other conditions com-bined was also significant, b 5 20.51, t 5 4.80, p 5.000003. Regarding effects on other aspects of visionquality, temporal projection caused leaders to craftvisionswith fewer values (p5 .041), which, as notedabove, is associated with greater vision quality, andmarginally increased the achievability of their vi-sions (p 5 .08).

We then tested how temporal projection impactsemployee inspiration once leaders have communi-cated their visions to their employees.We found thatbootstrapped estimates of the 95% confidence in-terval of the indirect effect (b5 .10) between leaders’use of mental time travel, the amount of image-ladenrhetoric in leaders’ visions, and their employees’level of inspiration excluded 0 [95%CI: .01, .29]. Theconfidence intervals for this indirect effect contin-ued to exclude zero when including the three otheraspects of vision quality as covariates. Finally, whenwe tested the effect of all four elements of visionquality simultaneously as mediators between tem-poral projection and employee inspiration, imagerywas the strongest mediator and the 95% confidenceinterval of the overall indirect effect (b 5 .14) ex-cluded 0 [95% CI: .01, .36].

SUPPLEMENTARY ANALYSES

There is consensus among linguists that individ-ual words are the most important ingredient foreliciting imagery because they serve as the “rawmaterial” that determines whether people arecapable of seeing a word in their mind’s eye(e.g., “yellow” versus “protocol”) (Guadagno et al.,2011). However, imagery is also shaped by how in-dividuals interpret higher-order components of lan-guage, including clauses and the entire statement asa whole (Pinker, 1995). Thus, we tested an alterna-tive measure of our dependent variable featuring

undergraduate raters. Two raters who were blind tothe study hypotheses were trained by assessing 50randomly selected visions of varying image strengthand quality, including several visions that werepretested to vary according to imagery and not anyother dimension. We then assessed whether theycorrectly identified statements that were more ver-sus less image-laden. They then independently rateda randomly selected subsample of 50 participants’visions at two levels of analysis: (1) distinct clauseswithin each vision statement (e.g., “cars drivingthrough the streets”), and (2) the entire vision state-ment as awhole. If thereweremultiple images acrossa statement, the raters assessed how well the differ-ent images complemented each other. We did notdiscuss the coding with the raters after the initialtraining, review their codes with them, ask them toalter codes, or retrain them to achieve higher re-liability. Rather, we analyzed only the results theysent us the first time.

We first assessed reliability at the clause level. Itwas possible to assess intrarater (within-person) re-liability because most visions had multiple clausesin which separate images could be embedded(e.g., “people driving their cars,” “commuters ridingthe subway”). Along these lines, there was accept-able reliability among separate clauses within eachindividual vision for both the first rater (a 5 .952)and the second rater (a 5 .766). We then assessedinterrater reliability. First,we assessed it for differentclauses within each statement (the clause level).Reliability was acceptable for the first (a 5 .767),second (a5 .670), and third (a5 .702) clauses; therewere not enough statements with more than threeclauses to perform analyses. We then assessedinterrater reliability for imagery of the vision state-ments as a whole. This level of reliability was alsoacceptable (a 5 .811). We then assessed predictivevalidity. Mental time travel boosted the overall im-age quality of leaders’visions for the first rater at boththe clause level, F(1,43) 5 4.717, p 5 .036, andstatement level, F(1,43)5 5.209, p5 .028, as well asfor the second rater at both the clause level, F(1,43)59.420, p5 .004, and statement level, F(1,43)5 7.323,p 5 .010.

GENERAL DISCUSSION

Given that leadership involves inspiring people toachieve a common purpose (Hemphill & Coons,1957; Locke, 1999), the ability to establish a sense ofpurpose is central to effective leadership. The resultsof our three experiments, featuring samples ranging


from British government officials to senior corporateexecutives, can redirect research on one of the mostimportant rhetorical tactics that a leader uses to es-tablish a common purpose—a vision of the future.

Upending an Assumption in Research onLeader Communication

Image-based rhetoric improves vision communi-cation (Masuda et al., 2010), yet leaders include littleof it when communicating visions of the future(Emrich et al., 2001). Consequently, employees oftenhave a blurry view of what their organizations aretrying to achieve (Carton et al., 2014). Given that vi-sions are communicated through verbal and writtenlanguage, it is tempting to assume that the remedy tothe blurry vision bias would involve a focus onselecting more appropriate words and phrases.Indeed, our review of 180 sources on vision com-munication (see Appendix B at blurryvisionbias.wordpress.com) suggests that the vast majority ofexperts assume leaders are best positioned to improvevision communication by carefully scrutinizing thewords they choose. Our findings upend this assump-tion. Leaders who averted their immediate attentionfrom the words they use and instead imagined what itwould be like to witness their organizations realizingtheir visions one day in the future (i.e., temporal pro-jection, or mental time travel) communicated visionsthat possessed greater imagerywhile not altering (and,in three cases, boosting) the other features of visionquality (achievability, specificity, and the expressionof values).

Incorporating Dual Cognitive Processing intoTheory on Leader Vision Communication

To explain why the blurry vision bias persists, aswell as approaches to correct it, we provided supportfor a theoretical framework that integrates pre-viously disconnected ideas from research on lead-ership, linguistics, and dual cognitive processing(Shondrick et al., 2010). In turn, our findings illu-minate both why temporal projection is effective forimproving vision communication (i.e., the psycho-logical process that explains its effectiveness) andwhen it is most effective (i.e., which leaders benefitmost from it).

Using dual cognitive processing to identify whytemporal projection boosts vision quality. Ourfindings suggest that an explicit focus on wordchoice causes leaders to overemphasize the abstractmeaning of words and leaves the experience-based

system of cognition deactivated. The experience-based system graphically simulates the observableworld and thuspositions leaders to craft image-basedrhetoric about the future. Even when leaders attuneto the properties of language responsible for trigger-ing mental images (verbs and nouns that describeobservable reality), they become so focused on lan-guage that only the language center of the brain isactivated (Tulving, 1972). Since the language centeris a component of meaning-based processing ratherthan experience-based processing, leaders do notadequately engage the experience-based system, and,in turn, do not transcend their natural inclination toformulateabstractvisions rather thanvividones, evenwhen they are explicitly instructed to consider thetypes of words that visually represent observable re-ality. In contrast, leaders who mentally projectthemselves to a moment in the future when their or-ganization’s vision has been achieved benefit fromactivation of the experience-based system. In-dividuals are prompted to “see” what the observableworld will look like when their vision is realized—such as facial expressions customers make as theyenjoy an event, or the look of a new technologicaldevice. Once individuals construct a “movie scene ofthe future” to “play” in their minds, they need to dolittle consciouswordsmithing tocommunicate amoreinspiring vision. Rather, they can increase the effec-tiveness of their vision communication by translatingthis mental movie scene into words. In doing so, theyare able to retrieve thewords needed to communicateavision that is vividwhile also strong in termsof otheraspects of vision quality (specificity, achievability,and the expression of values).

Accordingly, the most powerful method forredressing the blurry vision bias is a circuitousone—a path that detours in the system of cognitionthat is oftentimes neglected by scholars and practi-tioners. Although meaning-based processing is crit-ical because it allows individuals to use data-drivenand evidence-based decision protocols (Pfeffer &Sutton, 2006), as well as engage in sensemaking andhigher-order cognitive processing (Simon & Newell,1971), our research suggests that deliberate stepsneed to be taken to activate the experience-basedsystem. Since the future has not yet transpired, it isnot currently being experienced (as is the case withthe present), nor has it already been experienced (asis the case with the past). Accordingly, people in-herently do not contemplate the future with sensoryinformation. An active intervention (i.e., temporalprojection) thus helps leaders simulate what the fu-ture could look like and better communicate about it.




Using dual cognitive processing to identify whobenefits the most from temporal projection. Ourincorporation of dual cognitive processing alsocontributes to theory on leadership by illuminatingwhich leaders are most likely to benefit from psy-chologically projecting themselves to an importantmoment in the distant future. Leaders who are ef-ficient thinkers (i.e., are predisposed to make in-tuitive decisions) allow the imagery that theyconjure in the experience-based system as a resultof temporal projection to influence the languagethey use when communicating visions (Epsteinet al., 1996). In contrast, leaders who are inclined tothink complexly (i.e., analytical thinkers) do notbenefit as much from such an intervention. Their ce-rebral nature blocksmental imagery from influencingtheir rhetoric. Although analytical thinkers are oftenextolled for their capacity to systematically vet situ-ations and reach sound conclusions, our findings addto concerns raised in recent research related to thedownsides of analytical thinking—such as a de-creased ability to be decisive (Dalal & Brooks, 2013)and a reluctance to exploit the upsides of intuition(Dane & Pratt, 2007). In the case of vision communi-cation, leaders who are inclined to let their imagina-tion reign free are likely to be themost effective visioncommunicators.

Implications for Leader Communication inOther Domains

By establishing that current understanding ofleader communication is misalignedwith the natureof dual cognitive processing and then updating ourunderstanding of how these two areas of researchintersect, scholars can better understand how to in-corporate theory on dual cognitive processing intoresearch on other aspects of leader communication.One such topic involves feedback. In a meta-analysis of dozens of studies, Kluger and DeNisi(1996) found that individuals improve their per-formance when they are given concrete feedbackabout which task-specific behaviors they shouldenact (“please arrive to work at 9 am”) rather thangeneral feedback (“please be more conscien-tious”), yet leaders often provide general feedbackabout personal traits. To take corrective action,leaders can mentally simulate specific task be-haviors in the experience-based system prior togiving employees feedback. Similarly, when giv-ing task instructions leaders often know a task sowell that they fail to appreciate what it is like notto know it, leading them to communicate in terms

that an expert can understand (abstract jargon)rather than in terms a novice can understand(concrete detail related to learning new behaviors)(Keysar, Lin, & Barr, 2003). Leaders who engagetheir experience-based system by imagining whatit is like to perform a specific job may give bettertask instructions. Finally, our framework can in-form theory on culture. Leaderswho tell stories areoften more effective at communicating an organi-zation’s culture compared to those who onlycommunicate values (Schein, 1990). It is possiblethat leaders who use mental time travel to “jumpbackward” in time will be more likely to identifyvivid moments from the organization’s past thatthey can convey as stories to exemplify their or-ganization’s desired culture.

Empirical Contributions, Practical Contributions,and Limitations

We augmented our theoretical contributions bymaking a number of empirical contributions: (1) us-ing a method that allows for a blend of internal andexternal validity, (2) boosting internal validity fur-ther by controlling for confounds and directlyassessing mediating mechanisms, (3) demonstratingthat our theory works with several ecologically validsamples, (4) replicating the core effects multipletimes, (5) introducing a new approach to reviewingan extensive body of literature, and (6) testing fourdependent measures, such that we included twoseparate measures each for vision imagery andoverall vision quality—one that approximated theconstituent elements (the parts that make up thewhole) and one that approximated a holistic “ge-stalt” impression.

In terms of practical contributions, our paper canhelp close a gap between the existence of knowledge(image-laden visions have beneficial consequences)and the implementation of that knowledge (identi-fying the theoretical conditions that help leaderscraft and communicate image-based visions) be-cause we not only established the effectiveness oftemporal projection, but provided a recipe in themethods section for how it can be implemented. It isour hope that organizational leaders can use themethods in this paper to construct amental catalog ofimages and paint a vivid portrait of the future thatgalvanizes organizational members. Another practi-cal contribution relates to our finding that analyticalleaders struggle to harness the potential of temporalprojection. This lends credence to the idea thatorganizations may be best served by coupling an


inspirational visionarywith a cerebral strategist whomanages operational details—a dichotomy famouslyembodiedbySteve Jobs andTimCookatApple in themid-2000s.

There are several limitations with our paper. Al-though we found consistent results for moderatedmediation across two studies, we urge caution withrespect to the interpretation of these effects becauseof the nonnormal distributions of the mediator andmoderator, and that fact that these effects appear tobe driven by an extremely strong relationship be-tween the independent variable and the mediator(p , .0001), which in turn amplifies the moderatedeffect on the path between the mediator and the de-pendent variable. Although this meets the criteriafor a significant conditional indirect effect (Hayes,2013), the moderated effect is less powerful thanthe mediated effect. Another consideration is themental time travel manipulation. Although the word“company” was vetted, and its face validity de-termined, for the Brexit study, it may have been lessappropriate for participantswhoworked in agencies.A final limitation relates to the Brexit sample. Al-though we sampled leaders with large spans of con-trol and high-ranking positions in Studies 2 and 3,some participants in the Brexit study did not havehigh-ranking roles.

A number of future directions are likely to befruitful for scholars who seek to build on the currentfindings, as well as to address the limitations of ourstudies. Itmay beuseful to explorewhether temporalprojection generalizes to team leaders. For instance,we ran a supplementary experiment using the sameexperimental design as in study 3. Daytime and ex-ecutiveMasters of BusinessAdministration students(n 5 121) crafted visions for a team they oversaw.Linear mixed models established that those in themental time travel condition communicated visionswith more imagery, b 5 10.03, t 5 2.49, p 5 .01.Future research could also examine the role of nu-ances of vision communication thatwere outside thescope of this paper, including delivery, cadence, andfollower characteristics (Awamleh & Gardner, 1999;Stam et al., 2014). Additionally, scholars may ex-amine whether individuals can sustain their abilityto communicate high-quality visions long after en-gaging in temporal projection—a possibility thatseems likely due to the length of time that visualthoughts occupy memory (Tulving, 1972). Finally,scholars can explore tactics leaders use to furtherrefine their visions after they have engaged in tem-poral projection and converted their mental imageryto words.

CONCLUSION

Leaders can promote effective individual andcollective action by conveying a vivid sense of whatthe distant future could look like. Image-laden rhet-oric is a vehicle that enables leaders to achieve this,yet our findings suggest that leaders are inclined (andadvised) to engage in rhetorical tactics that causethem to communicate vague descriptions of the fu-ture. In contrast, temporal projection—a tactic thatimpels leaders to vividly imagine a real-life scenarioin the distant future and then translate it intowords—can help leaders craft and transmit high-quality visions. Temporal projection places leadersinto a mindset that exploits the often untapped re-source of the imagination, leaving them poised tocommunicate a verbal portrait that captures atten-tion and inspires action.

REFERENCES

Allen, P. A., Kaut, K. P., & Lord, R. R. 2008. Emotion andepisodic memory. In E. Dere, A. Easton, L. Nadel, &J. P. Huston (Eds.), Handbook of behavioral neuro-science, vol. 18: 115–132. Amsterdam, Netherlands:Elsevier Science.

Ashford, S. J., Wellman, N., Sully de Luque, M., DeStobbeleir, K. E., & Wollan, M. 2017. Two roads toeffectiveness: CEO feedback seeking, vision articula-tion, and firm performance. Journal of Organiza-tional Behavior, 39: 82–95.

Awamleh, R., & Gardner, W. L. 1999. Perceptions of leadercharisma and effectiveness: The effects of visioncontent, delivery, and organizational performance.The Leadership Quarterly, 10: 345–373.

Babin, L. A., & Burns, A. C. 1998. A modified scale for themeasurement of communication‐evoked mental im-agery. Psychology and Marketing, 15: 261–278.

Baetz, M. C. B., & Kenneth, C. 1998. The relationship be-tween mission statements and firm performance: Anexploratory study. Journal of Management Studies,35: 823–853.

Bass, B. M. 1990. From transactional to transformationalleadership: Learning to share the vision. Organiza-tional Dynamics, 18: 19–31.

Bator, R., & Cialdini, R. 2000. The application of persua-sion theory to the development of effective pro-environmental public service announcements. TheJournal of Social Issues, 56: 527–542.

Bell, A., & Jones, K. 2015. Explaining fixed effects: Randomeffects modeling of time-series cross-sectional andpanel data. Political Science Research andMethods,3: 133–153.


Berg, J. M. 2014. The primal mark: How the beginningshapes the end in the development of creative ideas.Organizational Behavior and Human DecisionProcesses, 125: 1–17.

Breen, R. 1996. Regression models: Censored, sampleselected, or truncated data. Atlanta, GA: Sage.

Buhrmester, M., Kwang, T., & Gosling, S. D. 2011. Ama-zon’s Mechanical Turk: A new source of inexpensive,yet high-quality, data?Perspectives onPsychologicalScience, 6: 3–5.

Button, K. S., Ioannidis, J. P. A., Mokrysz, C., Nosek, B. A.,Flint, J., Robinson, E. S. J., &Munafo,M. R. 2013. Powerfailure: Why small sample size undermines the re-liability of neuroscience. Nature Reviews. Neurosci-ence, 14: 365–376.

Bulley, A., Henry, J., & Suddendorf, T. 2016. Prospectionand thepresentmoment: The role of episodic foresightin intertemporal choices between immediate anddelayed rewards. Review of General Psychology, 20:29–47.

Cable,D.M.,Gino, F., &Staats, B.R. 2013.Breaking them inor eliciting their best? Reframing socialization aroundnewcomers’ authentic self-expression. Administra-tive Science Quarterly, 58: 1–36.

Cacioppo, J. T., Petty, R. E., Feinstein, J. A., & Jarvis,W. B. G.1996. Dispositional differences in cognitive motiva-tion: The life and times of individuals varying in needfor cognition. Psychological Bulletin, 119: 197–253.

Carton, A. M. 2018. “I’m not mopping the floors, I’m put-ting a man on the moon”: How NASA leaders en-hanced the meaningfulness of work by changing themeaning of work.Administrative ScienceQuarterly,63: 323–369.

Carton, A. M., Murphy, C., & Clark, J. R. 2014. A (blurry)vision of the future: How leader rhetoric about ulti-mate goals influences performance. Academy ofManagement Journal, 57: 1544–1570.

Collins, J. C., & Porras, J. I. 1994. Built to last. New York,NY: Harper Collins.

Crossley, S., Salsbury, T., & McNamara, D. 2009. Measur-ing L2 lexical growthusing hypernymic relationships.Language Learning, 59: 307–334.

Dalal, R. S., & Brooks, M. E. 2013. Individual differences indecision-making skill and style. In S. Highhouse, R. S.Dalal & E. Salas (Eds.), Judgment and decision mak-ing at work: 80–101. New York, NY: Routledge.

Dane, E., & Pratt, M. G. 2007. Exploring intuition and itsrole in managerial decision making. Academy ofManagement Review, 32: 33–54.

Edison Investment Research, 2014. Taking the next step.Retrieved fromhttps://www.edisoninvestmentresearch.com/?ACT518&ID512465&LANG5.AccessedOctober15, 2018.

Emrich, C. G., Brower, H. H., Feldman, J. M., & Garland, H.2001. Images in words: Presidential rhetoric, cha-risma, and greatness. Administrative Science Quar-terly, 46: 527–557.

Epstein, S., Pacini, R., Denes-Raj, V., & Heier, H. 1996. In-dividual differences in intuitive–experiential andanalytical–rational thinking styles. Journal of Per-sonality and Social Psychology, 71: 390–405.

Eyal, T., Liberman, N., & Trope, Y. 2008. Judging near anddistant virtue and vice. Journal of Experimental So-cial Psychology, 44: 1204–1209.

Fletcher, P., Frith,C.,Baker, S., Shallice,T., Frackowiak,R.,&Dolan, R. 1995. The mind’s eye—Precuneus activationin memory-related imagery.NeuroImage, 2: 195–200.

Gardner, W. L., & Schermerhorn, J. J. R. 2004. Unleashingindividual potential: Performance gains through pos-itive organizational behavior and authentic leader-ship. Organizational Dynamics, 33: 270–281.

Graesser, A. C., McNamara, D. S., Louwerse, M. M., & Cai,Z. 2004. Coh-Metrix: Analysis of text on cohesion andlanguage. Behavior Research Methods, Instruments,& Computers, 36: 193–202.

Grant, A. M. 2008. Does intrinsic motivation fuel the pro-social fire? Motivational synergy in predicting per-sistence, performance, and productivity. The Journalof Applied Psychology, 93: 48–58.

Grant, A. M., Campbell, E. M., Chen, G., Cottone, K.,Lapedis, D., & Lee, K. 2007. Impact and the art of mo-tivation maintenance: The effects of contact withbeneficiaries on persistence behavior. Organiza-tional Behavior and Human Decision Processes,103: 53–67.

Grant, A. M., & Hofmann, D. A. 2011. Outsourcing in-spiration: The performance effects of ideologicalmessages from leaders and beneficiaries. Organiza-tional Behavior and Human Decision Processes,116: 173–187.

Greer, L. L., Homan, A. C., De Hoogh, A. H., & Den Hartog,D. N. 2012. Tainted visions: The effect of visionaryleader behaviors and leader categorization tendencieson the financialperformanceof ethnicallydiverse teams.The Journal of Applied Psychology, 97: 203–213.

Guadagno,R. E., Rhoads,K. L., &Sagarin, B. J. 2011. Figuralvividness and persuasion: Capturing the “elusive”vividness effect. Personality and Social PsychologyBulletin, 37: 626–638.

Hansen, E. G. 2010. Responsible leadership systems: Anempirical anlaysis of integrating corporate re-sponsibility into leadership systems. Wiesbaden,Germany: Gabler.

Hayes,A. F. 2013. Introduction tomediation,moderation,and conditional process analysis: A regression-based approach. New York, NY: Guilford Press.


https://www.edisoninvestmentresearch.com/?ACT&tnqh_x3d;18&tnqh_x26;ID&tnqh_x3d;12465&tnqh_x26;LANG&tnqh_x3d;




Heath, C., & Heath, D. 2010. Switch: How to change whenchange is hard. New York, NY: Broadway Books.

Hemphill, J. K., & Coons, A. E. 1957. Development of theleader behavior description questionnaire. In R. M.Stodgill & A. E. Coons (Eds.), Leader behavior: Itsdescription andmeasurement: 6–38. Columbus, OH:Bureau of Business Research, Ohio State University.

Kahneman, D. 2011. Thinking, fast and slow. New York,NY: Farrar, Straus and Giroux.

Keysar, B., Lin, S., & Barr, D. J. 2003. Limits on theory ofmind use in adults. Cognition, 89: 25–41.

Kluger, A. N., & DeNisi, A. 1996. The effects of feedbackinterventions on performance: a historical review,a meta-analysis, and a preliminary feedback inter-vention theory.Psychological Bulletin, 119: 254–284.

Kotter, J. P. 1996. Leading change. Cambridge, MA:Harvard Business School Press.

Locke, E. A. 1999. The essence of leadership: The fourkeys to leading successfully. New York, NY:Lexington.

Locke, E. A., & Latham, G. P. 1990.A theory of goal settingand task performance. Englewood Cliffs, NJ: Pren-tice-Hall.

Lucas,B. J.,&Nordgren,L.F. 2015.Peopleunderestimate thevalue of persistence for creative performance. Journalof Personality and Social Psychology, 109: 232–243.

Magee, J. C., Milliken, F. J., & Lurie, A. R. 2010. Powerdifferences in the construal of a crisis: The immediateaftermath of September 11, 2001. Personality andSocial Psychology Bulletin, 36: 354–370.

Masuda,A.D., Kane, T. D., Shoptaugh, C. F., &Minor, K.A.2010. The role of a vivid and challenging personalvision in goal hierarchies.The Journal of Psychology,144: 221–242.

Miller, C. C., & Cardinal, L. B. 1994. Strategic planning andfirm performance: A synthesis of more than two de-cades of research.Academy ofManagement Journal,37: 1649–1665.

Mount, S. 2010. The “I have a dream” speech. USCon-stitution.net.Retrievedfromhttp://www.usconstitution.net/dream.html. Accessed October 15, 2018.

Munafo, M. R., Nosek, B. A., Bishop, D. V., Button, K. S.,Chambers, C. D., du Sert, N. P., Simonsohn, U.,Wagenmakers, E.-J., Ware, J. J., & Ioannidis, J. P. 2017.Amanifesto for reproducible science.Nature HumanBehaviour, 1: 0021.

Nanus, B. 1992. Visionary leadership: Creating a com-pelling sense of direction for your organization. SanFrancisco, CA: Jossey-Bass.

Nisbett, R., & Ross, L. 1980. Human inference: Strategiesand shortcomings of social judgment. EnglewoodCliffs, NJ: Prentice-Hall.

Nordgren, L. F., & McDonnell, M.-H. M. 2011. The scope-severity paradox.Social Psychological&PersonalityScience, 2: 97–102.

O’Reilly, C. A., Chatman, J., & Caldwell, D. F. 1991. Peopleand organizational culture: A profile comparison ap-proach to assessing person-organization fit.Academyof Management Journal, 34: 487–516.

Paolacci, G., Chandler, J., & Ipeirotis, P. G. 2010. Runningexperiments on Amazon Mechanical Turk. Judgmentand Decision Making, 5: 411–419.

Pfeffer, J., & Sutton, R. I. 2006. Evidence-based manage-ment. Harvard Business Review, 84: 62–74.

Pinker, S. 1995. The language instinct: The new scienceof languageandmind.NewYork,NY:HarperCollins.

Pratt, M. G., & Ashforth, B. E. 2003. Fostering meaning-fulness in working and at work. In K. S. Cameron, J. E.Dutton, & R. E. Quinn (Eds.), Positive organizationalscholarship: Foundations of a new discipline:309–327. San Francisco, CA: Berrett-Koehler.

Rafferty, A. E. and Griffin, M. A. 2004. Dimensions oftransformational leadership: Conceptual and empiri-cal extensions. The Leadership Quarterly, 15:329–354.

Rosch, E. 1978. Principles of categorization. In E. Rosch &B. B. Lloyd (Eds.), Cognition and categorization:27–48. Hillsdale, NJ: Erlbaum.

Rosso, B. D., Dekas, K.H., &Wrzesniewski, A. 2010. On themeaning of work: A theoretical integration and re-view. Research in Organizational Behavior, 30:91–127.

Rucker, D. D., Galinsky, A. D., & Dubois, D. 2012. Powerand consumer behavior: How power shapes who andwhat consumers value. Journal of Consumer Psy-chology, 22: 352–368.

Schacter, D. L., Addis, D. R., & Buckner, R. L. 2007. Re-membering thepast to imagine the future: theprospectivebrain.Nature Reviews. Neuroscience, 8: 657–661.

Schein, E. H. 1990. Organizational culture. AmericanPsychologist, 45: 109–119.

Schwarz, N., Bless, H., Strack, F., Klumpp, G., Rittenauer-Schatka, H., & Simons, A. 1991. Ease of retrieval asinformation: Another look at the availability heuristic.Journal of Personality and Social Psychology, 61:195–202.

Shondrick, S. J., Dinh, J. E., & Lord, R. G. 2010. De-velopments in implicit leadership theory and cognitivescience: Applications to improving measurement andunderstanding alternatives to hierarchical leadership.The Leadership Quarterly, 21: 959–978.

Simmons, J. P., Nelson, L. D., & Simonsohn, U. 2011.False-positive psychology: Undisclosed flexibilityin data collection and analysis allows presenting


http://www.usconstitution.net/dream.html

http://www.usconstitution.net/dream.html

anything as significant. Psychological Science, 22:1359–1366.

Simon, H. A., & Newell, A. 1971. Human problem solving:The state of the theory in 1970. The American Psy-chologist, 26: 145–159.

Small, D. A., Loewenstein, G., & Slovic, P. 2007. Sympathyand callousness: The impact of deliberative thoughton donations to identifiable and statistical victims.Organizational Behavior and Human DecisionProcesses, 102: 143–153.

Smith, P.K., &Trope,Y. 2006.You focuson the forestwhenyou’re in charge of the trees: Power priming and ab-stract information processing. Journal of Personalityand Social Psychology, 90: 578.

Smith, S. M., & Shaffer, D. R. 2000. Vividness can un-dermine or enhance message processing: The moder-ating role of vividness congruency. Personality andSocial Psychology Bulletin, 26: 769–779.

Stanton, E. C., Gordon, A. D., & Anthony, S. B. 1997. TheSelected Papers of Elizabeth Cady Stanton andSusan B. Anthony: When Clowns Make Laws forQueens, 1880-1887. Rutgers University Press.

Stam, D., Lord, R. G., van Knippenberg, D., & Wisse, B.2014. An image of who we might become: Visioncommunication, possible selves, and vision pursuit.Organization Science, 25: 1172–1194.

Stephens-Davidowitz, S. 2017. Everybody lies: Bigdata, new data, and what the internet can tell usabout who we really are. New York, NY:HarperCollins.

Suddendorf, T., Addis, D. R., & Corballis, M. C. 2009.Mental time travel and the shaping of the humanmind. Philosophical Transactions of the Royal

Society of London. Series B, Biological Sciences,364: 1317–1324.

Thrash, T. M., & Elliot, A. J. 2003. Inspiration as a psycho-logical construct. Journal of Personality and SocialPsychology, 84: 871–889.

Trope, Y., & Liberman, N. 2003. Temporal construal. Psy-chological Review, 110: 403–421.

Tulving, E. 1972. Episodic and semantic memory. In E.Tulving & W. Donaldson (Eds.), Organization ofmemory: 381–403. New York, NY: Academic Press.

van Knippenberg, D., & Sitkin, S. B. 2013. A critical as-sessment of charismatic—Transformational leader-ship research: Back to the drawing board? TheAcademy of Management Annals, 7: 1–60.

Waytz, A., Hershfield, H. E., & Tamir, D. I. 2015. Mentalsimulation and meaning in life. Journal of Personal-ity and Social Psychology, 108: 336.

Andrew M. Carton ([email protected]) is an assis-tant professor of management at The Wharton School at theUniversity of Pennsylvania. He obtained his PhD from DukeUniversity. He studies how leaders establish a shared sense ofpurpose, as well as how they manage intergroup conflict.

Brian J. Lucas ([email protected]) is an assistantprofessor of organizational behavior at the School of In-dustrial and Labor Relations at Cornell University. He re-ceived his PhD from Northwestern University’s KelloggSchool of Management. He studies the psychology of cre-ativity, as well as ethics and morality.


mailto:[email protected]

mailto:[email protected]

How Can Leaders Overcome the Blurry Vision Bias ......BRIAN J. LUCAS Cornell University Evidence...

Documents

Transcript of How Can Leaders Overcome the Blurry Vision Bias ......BRIAN J. LUCAS Cornell University Evidence...