Using Linguistic Analysis to Identify High Performing Teams Mary T. Dzindolet & Linda G. Pierce...
-
Upload
silas-fields -
Category
Documents
-
view
217 -
download
2
Transcript of Using Linguistic Analysis to Identify High Performing Teams Mary T. Dzindolet & Linda G. Pierce...
Using Linguistic Analysis to Using Linguistic Analysis to Identify High Performing TeamsIdentify High Performing Teams
Mary T. Dzindolet & Linda G. Pierce
Cameron University Army Research Laboratory
2006 CCRTS2006 CCRTSTHE STATE OF THE ART AND THE STATE OF THE PRACTICETHE STATE OF THE ART AND THE STATE OF THE PRACTICE
Demands on Military TeamsDemands on Military Teams
• Perform a wide variety of tasks– Peace-keeping– War
• Ever-changing – Team members– Situation– Leadership
• High threat
PurposePurpose
This presentation will explore the usefulness of one technological tool, the Linguistic Inquiry Word Count (LIWC), in
identifying high-performing teams.
LIWC VariablesLIWC Variables
Category (Examples)• pronouns (I, me, we, you) • positive emotions (happy, pride, good)• negative emotions (hate, afraid, sad)• insight (think, know, consider)• time (past, present, future)• communication (talk, share, converse)• anxiety words (nervous, afraid, tense)
LIWC analyzes text word-by-word and categorizes the text into 74 different linguistic dimensions
Prior ResearchPrior Research
• Pennebaker, Mehl, and Niederhoffer (2003) used the LIWC in dyadic social interactions to understand relationships
• We have extended this work to explore the usefulness of the LIWC in identifying high-performing teams
Predicting Group PerformancePredicting Group Performance• Groups may differ in many important ways
– Size (number of members)– Degree to which members are stratified (hierarchical)– Degree to which members exercise control over the behavior of other
members– Degree of participation expected, permitted, or demanded of members– Ease of access to membership in the group and ease with which member
can leave or be expelled from the group
Predicting Group PerformancePredicting Group Performance• Groups may differ in many important ways
– Degree of stability of the group over time and the continuity of its members over time
– Degree to which group members relate to one another intimately vs formally– Degree to which the group is subdivided into smaller groups or cliques, and
the extent to which such cliques are in conflict with one another– Degree of homogeneity among group members– Type of Task***
G E N E R AT E
E X E C U T E
N E G O T IAT E
C H O O S E
o f v iew p o in t
G enera tin g id ea s G enera tin g p lans
w/no r igh t an sw
er
o f in terest
Dec id ing is sue s R eso lv in g con f licts R e so lv in g con f lic
ts
Res
o lv i
ng c
o nf l
i ct s
Ty pe 2 :c reativ ity task s
Ty pe 1 :p lann in g ta sks
Ty pe 3 :in tellec tiv e ta sk s
Ty pe 8 :pe rform an ce/
psycho -m o to r task s
Ty pe 4 :dec isio n-m ak in g
task s
Ty pe 7 :con te sts/b attle s/
com p etitive ta sks
Ty pe 6 :m ixed-m otiv e
task s
Ty pe 5 :cog nitive
con flic ts ta sk s
w/c
orr e
ct a
n sw
ers
o f p
ower
Solv
ing
p ro b
lem
s E xecu tin g
per fo rmanc e tas k s
Con
flic
tC
oope
ratio
n
C on cep tu al B eh av iora l
M cG rath ’s C ircu m p lex o f G ro u p Ta sk s
G E N E R AT E
E X E C U T E
N E G O T IAT E
C H O O S E
o f v iew p o in t
G enera ting id e as G en era tin g p la ns
w/n o r igh t an sw
er
of in te re st
Dec id in g issu es R e so lv in g co n f licts R e so lv in g con f lic
ts
Re s
o lvi
ng c
o nf l
i ct s
Typ e 2 :c rea tiv ity ta sk s
Typ e 1 :p lan n in g ta sk s
Typ e 3 :in te llec tiv e ta sks
Typ e 8 :p erfo rm an ce /
p sych o -m o tor tasks
Typ e 4 :d ec is io n-m akin g
tasks
Typ e 7 :con te sts /b a ttle s/
com pe titiv e tasks
Typ e 6 :m ix ed -m otive
ta sks
Typ e 5 :cog n itive
con flic ts tasks
w/c
orre
c t a
nsw
ers
o f p
owe r
Sol
v ing
pro
b le m
s Exec ut in g
p e r fo rman ce task
s
Con
flic
tC
oope
ratio
n
C on cep tu a l B ehav ioral
M cG ra th ’s C ircu m p lex o f G ro u p Task s
Typ e 2 :c rea tiv ity ta sk s
Predicting Group PerformancePredicting Group PerformanceTask Type 2: Generating IdeasTask Type 2: Generating Ideas
Brainstorming TasksBrainstorming Tasks
• Can linguistic analysis predict performance on a brainstorming task?
• We examined correlations between number of ideas generated and LIWC variables for nine studies performed either at Cameron University or the University of Texas at Arlington
• Time to Brainstorm – 5, 10, 20, or 45 minutes
• Group Cohesiveness• Research Location
– Cameron University, University of Texas at Arlington, Walmart, or City National Bank
• Group Size – dyad, triad, or quad
• Communication Medium – face-to-face, distributed, or groupshareware
• Brainstorming Problem
Study DifferencesStudy Differences
Brainstorming RulesBrainstorming Rules
• Criticism is ruled out. Adverse judgement of ideas must be withheld.
• Freewheeling is welcome, the wilder the idea, the better.
• Quantity is wanted. Come up with as many as you can.
• Combination and improvement are sought. Do not be afraid to combine and improve on ideas.
Pronoun Use
Total PronounsTotal Pronouns(I, our, they, you, we)
Study
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce
* *
*
* *
Thumbs Improve Univ Ecology Univ Teens
First Person Singular PronounsFirst Person Singular PronounsI, me, my
Study
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce
* *
* **
Thumbs Improve Univ Ecology Univ Teens
First Person Plural PronounsFirst Person Plural Pronounswe, our, us
Study
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce
Second Person PronounsSecond Person Pronounsyou, your, y’all
Study
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce
Third Person PronounsThird Person Pronounshe, she, they
Study
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce
Why Is Use of First Person Pronouns Why Is Use of First Person Pronouns Related to Poor Performance?Related to Poor Performance?
• Self-focus rather than other-focus
References to Other PeopleReferences to Other Peoplethem, you, anyone, everybody, someone
Study
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce
Why Is Use of First Person Pronouns Why Is Use of First Person Pronouns Related to Poor Performance?Related to Poor Performance?
• Self-focus rather than other-focus
• Subordinate status of some group members
Tentative WordsTentative Wordsmaybe, perhaps, depending
Study
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce
Certainty WordsCertainty Wordsclearly, always, confidentlyclearly, always, confidently
Study
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce
Why Is Use of First Person Pronouns Why Is Use of First Person Pronouns Related to Poor Performance?Related to Poor Performance?
• Self-focus rather than other-focus
• Subordinate status of some group members
• Defending own views; lack of supportive group environment
NegateNegateno, never, not
Study
-0.8-0.7-0.6-0.5-0.4-0.3-0.2-0.1
00.1
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce
AssentAssentyes, O.K., alright, agree
Study
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce
*
Cognitive Complexity
Words with Six or More LettersWords with Six or More Letters
Study
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce
* *
Exclusive WordsExclusive Wordsbut, except, without
Study
-0.8
-0.7
-0.6
-0.5
-0.4
-0.3
-0.2
-0.1
0
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce
* * * *
Inclusive WordsInclusive Wordstogether, with, also
Study
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce
**
ArticlesArticlesa, an, the
Study
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce
PrepositionsPrepositionsto, for, at
Study
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce
NumbersNumbers
Study
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce
Causal WordsCausal Wordsbecause, since, basis
Study
-0.8
-0.6
-0.4
-0.2
0
0.2
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce
**
Emotions
Positive Emotion WordsPositive Emotion Wordshappy, pretty, good
Study
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce
*
*
Words Expressing Positive FeelingsWords Expressing Positive Feelingscare, encourage, enjoy
Study
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce
*
*
Optimistic WordsOptimistic Wordshope, best, win
Study
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce *
Negative Emotion WordsNegative Emotion Wordshate, worthless, ugly
Study
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce
**
Anxiety WordsAnxiety Wordsnervous, scared, anxious
Study
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce
Words Expressing AngerWords Expressing Angerjerk, kill, annoy
Study
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce
Words Expressing SadnessWords Expressing Sadnesssad, upset, suffer
Study
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce
*
Cognitive Processes
Cognitive MechanismsCognitive Mechanismsquestioning, acknowledge, inform
Study
-0.8-0.7-0.6-0.5-0.4-0.3-0.2-0.1
00.1
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce
* *
Insight WordsInsight Wordsthink, know, believe
Study
-0.8-0.7-0.6-0.5-0.4-0.3-0.2-0.1
00.1
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce
**
**
Discrepancy WordsDiscrepancy Wordsshould, ought, could
Study
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce
*
Social Processes
Social MechanismsSocial Mechanismsfriend, phone, gossip, group
Study
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce
*
Communication WordsCommunication Wordstalk, ask, chat, counsel
Study
-0.8-0.7-0.6-0.5-0.4-0.3-0.2-0.1
00.1
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce
*
Time
Past TensePast Tense
Study
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce
*
*
* *
Present TensePresent Tense
Study
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce
* * *
Future TenseFuture Tense
Study
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
1 2 3 4 5 6 7 8 9 10
* p<.05
co
rre
lati
on
wit
h p
erfo
rma
nce
Conclusion Conclusion Idea Generation TasksIdea Generation Tasks
• High-performing groups tend to– Avoid first person pronouns possibly due to group
support (few negations)– Use more cognitively complex language (avoid
exclusive and inclusive words and use words with six or more letters)
– Avoid communication words– Avoid words indicating cognitive processes including
causal words – Avoid negative emotion words– Avoid the present tense
• This pattern existed across many studies of groups that differed in size, communication medium, problem, location, and prior knowledge of one another
Predicting Group PerformancePredicting Group PerformanceTask Type 3: Solving Problems with Correct AnswersTask Type 3: Solving Problems with Correct Answers
Desert Survival ProblemDesert Survival Problem
Rank order 15 items in terms of utility for desert survival
Responses are compared with that of an expert
G E N E R AT E
E X E C U T E
N E G O T IAT E
C H O O S E
o f v iew p o in t
G en e ra t in g id ea s G en era t ing p la ns
w/ n
o r ig h t an swer
of in teres t
Dec id in g is su es R e so lv in g co n f lic ts R e so lv in g co n fl ic
ts
Re s
o lv i
ng c
onfl
i ct s
Ty p e 2 :c re a tiv i ty task s
Ty p e 1 :p la n n in g ta sk s
Ty p e 3 :in te l le c t iv e ta sk s
Ty p e 8 :p e rfo rm a n c e /
p sy c h o -m o to r ta sk s
Ty p e 4 :d e c isio n -m a k in g
ta sk s
Ty p e 7 :c o n te s ts /b a tt le s /
c o m p e ti tiv e ta sk s
Ty p e 6 :m ix e d -m o tiv e
ta sk s
Ty p e 5 :c o g n itiv e
c o n flic ts ta sk s
w/c
orre
c t a
n sw
e rs
o f p
ower
Sol v
ing
pro b
lem
s Ex ecu t ing
p e r forma n c e ta sk s
Con
flic
tC
oope
ratio
n
C o n c e p tu a l B e h a v io ra l
M cG ra th ’s C ircu m p lex o f G rou p Task s
Ty p e 3 :in te l le c t iv e ta sk s
Survival ProblemSurvival Problem
Survival ProblemSurvival Problem
• flashlight (4-battery size)
• jackknife
• sectional air map of the area
• plastic raincoat (large size)
• magnetic compass
• compress kit with gauze
• .45 caliber pistol (loaded)
• parachute (red and white)
• bottle of salt tablets (1000 tablets)
• 1 quart of water per person• book entitled, Edible Animals
of the Desert• 1 top coat per person• pair of sunglasses per person• cosmetic mirror• 2 quarts (2 liters) of 180 proof
Vodka
Pronoun UsePronoun Use
-1-0.8-0.6-0.4-0.2
00.20.40.60.8
Total FirstSingular
FirstPlural
Second Third
Dyad Triad * p<.05
corr
elat
ion
wit
h p
erfo
rman
ce
Cognitive ComplexityCognitive Complexity
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
6+ Causal Negate Excl Incl
Dyad Triad * p<.05
corr
elat
ion
wit
h p
erfo
rman
ce
Cognitive ComplexityCognitive ComplexityConcrete Words
-1-0.8-0.6-0.4-0.2
00.20.40.60.8
1
Articles Prepositions Numbers
Dyad Triad * p<.05
corr
elat
ion
wit
h p
erfo
rman
ce *
* *
Words Expressing Emotion
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
Affect PosEmo
PosFeel
NegEmot
Anger Sad
Dyad Triad
* p<.05
corr
elat
ion
wit
h p
erfo
rman
ce
Cognitive & Social Processes
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
Cog Mech Insight Discrep Social Comm
Dyad Triad
* p<.05
corr
elat
ion
wit
h p
erfo
rman
ce
Tense
-1-0.8-0.6-0.4-0.2
00.20.40.60.8
1
Past Present Future
Dyad Triad* p<.05
corr
elat
ion
wit
h p
erfo
rman
ce
ConclusionConclusionDesert Survival TaskDesert Survival Task
High performing dyads• Avoid prepositions and use
numbers (both indicators of concrete rather than abstract thought)
• Though not statistically significant, tend to – Avoid pronouns– Avoid negations– Use exclusive words– Use words with six or more
letters– Avoid discrepancy words– Avoid communication words– Avoid words expressing
emotion– Avoid present tense– Use the past tense
High performing groups of three
• Avoid prepositions• Though not statistically
significant, tend to– Use third person
pronouns– Express emotions– Use the past tense
Predicting Group PerformancePredicting Group PerformanceTask Type 3: Solving Problems with Correct AnswersTask Type 3: Solving Problems with Correct Answers
Student Government TaskStudent Government Task
Choose the best student government candidate using the characteristics of each candidates (Candidate A, B, and C)
Some information given to each group member is unique; other pieces of information are given to all group members
G E N E R AT E
E X E C U T E
N E G O T IAT E
C H O O S E
o f v iew p o in t
G en e ra t in g id ea s G en era t ing p la ns
w/ n
o r ig h t an swer
of in teres t
Dec id in g is sues R e so lv in g co n f lic ts R e so lv in g co n fl ic
ts
Re s
o lv i
ng c
onfl
i ct s
Ty p e 2 :c re a tiv i ty task s
Ty p e 1 :p la n n in g ta sk s
Ty p e 3 :in te l le c t iv e ta sk s
Ty p e 8 :p e rfo rm a n c e /
p sy c h o -m o to r ta sk s
Ty p e 4 :d e c isio n -m a k in g
ta sk s
Ty p e 7 :c o n te s ts /b a tt le s /
c o m p e ti tiv e ta sk s
Ty p e 6 :m ix e d -m o tiv e
ta sk s
Ty p e 5 :c o g n itiv e
c o n flic ts ta sk s
w/c
orre
c t a
n sw
e rs
o f p
ower
Sol v
ing
pro b
lem
s Ex ecu t ing
p e r forma n c e ta sk
s
Con
flic
tC
oope
ratio
n
C o n c e p tu a l B e h a v io ra l
M cG ra th ’s C ircu m p lex o f G rou p Task s
Ty p e 3 :in te l le c t iv e ta sk s
Pronoun UsePronoun Use
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
Total FirstSingular
FirstPlural
Second Third
Dyad Group * p<.05
*
Cor
rela
tion
with
Tim
e S
pent
on
Sh
ared
Sh
ared
Inf
o
Pronoun UsePronoun Use
-0.4
-0.2
0
0.2
0.4
0.6
0.8
Total FirstSingular
FirstPlural
Second Third
Dyad Group * p<.05
* *
Cor
rela
tion
with
Tim
e S
pent
on
Un
iqu
e In
fo
Cognitive ComplexityCognitive Complexity
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
6+ Causal Negate Excl Incl
Dyad Group * p<.05
* *
Cor
rela
tion
with
Tim
e S
pent
on
Sh
ared
Sh
ared
Inf
o
Cognitive ComplexityCognitive Complexity
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
6+ Causal Negate Excl Incl
Dyad Group * p<.05Cor
rela
tion
with
Tim
e S
pent
on
Un
iqu
e In
fo
* *
Cognitive ComplexityCognitive ComplexityConcrete Words
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
Articles Prepositions Numbers
Dyad Group * p<.05Cor
rela
tion
with
Tim
e sp
ent
on S
har
ed I
nfo
*
Cognitive ComplexityCognitive ComplexityConcrete Words
-0.2
0
0.2
0.4
0.6
Articles Prepositions Numbers
Dyad Group * p<.05 Cor
rela
tion
with
Tim
e S
pent
on
Un
iqu
e In
fo
Words Expressing Emotion
-0.4
-0.2
0
0.2
0.4
0.6
0.8
Affect PosEmo
PosFeel
NegEmot
Anger Sad
Dyad Group
* p<.05Cor
rela
tion
with
Tim
e S
pent
on
Sh
ared
Inf
o
* *
Words Expressing Emotion
-0.4
-0.2
0
0.2
0.4
Affect PosEmo
PosFeel
NegEmot
Anger Sad
Dyad Group
* p<.05Cor
rela
tion
with
Tim
e S
pent
on
Un
iqu
e In
fo
Cognitive & Social Processes
0
0.2
0.4
0.6
0.8
Cog Mech Insight Discrep Social Comm
Dyad Group
* p<.05Cor
rela
tion
with
Tim
e S
pent
on
Sh
ared
Inf
o
*
Cognitive & Social Processes
-0.4
-0.2
0
0.2
0.4
0.6
Cog Mech Insight Discrep Social Comm
Dyad Group
* p<.05Cor
rela
tion
with
Tim
e S
pent
on
Un
iqu
e In
fo
Tense
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
Past Present Future
Dyad Group * p<.05Cor
rela
tion
with
Tim
e S
pent
on
Sh
ared
Inf
o
*
* * *
Tense
-0.4
-0.2
0
0.2
0.4
Past Present Future
Dyad Group* p<.05
Cor
rela
tion
with
Tim
e S
pent
on
Un
iqu
e In
fo
ConclusionsConclusionsHidden Profile TaskHidden Profile Task
• Dyads using the efficient strategy– Use causal words
• Dyads using the inefficient strategy– Use causal word– Use second person
pronouns– Use exclusive words– Use prepositions– Use social words– Use present and future
tenses
• Groups of four using the efficient strategy– Use causal words– Use first person singular
and second person pronouns
• Groups of four using the inefficient strategy– Express emotions—
especially negative emotions
– Use the present tense– Avoid the past tense
Predicting Group PerformancePredicting Group PerformanceTask Type 4: Deciding Issues with No Task Type 4: Deciding Issues with No
Correct AnswerCorrect AnswerGroup Polarization TaskGroup Polarization Task
• Procedure– individual (no time limit) – group (30 minutes)– individual (no time limit)
• Indicate the lowest probability of succeeding that was acceptable for 12 different scenarios
“An electrical engineer may stick with his present job at a modest but adequate salary, or may take a new job offering considerably more money but no long-term security.”
G E N E R AT E
E X E C U T E
N E G O T IAT E
C H O O S E
o f v iew p o in t
G enera tin g id eas G en era tin g p lan s
w/no r ig h t an sw
er
o f in te rest
Dec id ing issu es R e so lv in g con f licts R e so lv in g con f lic
ts
Re s
o lv i
ng c
o nfl
i cts
Typ e 2 :c rea tiv ity ta sks
Typ e 1 :p lann in g ta sk s
Typ e 3 :in tellec tiv e task s
Typ e 8 :pe rform ance /
p sycho -m o to r ta sks
Typ e 4 :dec isio n-m ak ing
task s
Typ e 7 :con te sts /b a ttles/
com p e titiv e ta sk s
Typ e 6 :m ixed-m otive
task s
Typ e 5 :cog nitiv e
con flic ts task s
w/c
orre
c t a
nsw
ers
o f pow
er
Sol
vin g
pro
b lem
s Execu tin g
p er fo rma nce ta sk
s
Con
flic
tC
oope
ratio
n
C on cep tu a l B eh av ioral
M cG ra th ’s C ircu m p lex o f G ro u p Ta sk s
Typ e 4 :dec isio n-m ak ing
task s
Pronoun UsePronoun Use
-0.4-0.3-0.2-0.1
00.10.20.30.40.5
Total FirstSingular
FirstPlural
Second Third
FtF Dayd FtF Group Dist Group* p<.05
corr
elat
ion
wit
h a
gre
emen
t w
ith
gro
up
Cognitive ComplexityCognitive Complexity
-0.4
-0.2
0
0.2
0.4
0.6
6+ Causal Negate Excl Incl
FtF Dyad FtF Group Dist Group * p<.05corr
elat
ion
wit
h a
gre
emen
t w
ith
gro
up
Cognitive ComplexityCognitive ComplexityConcrete Words
-0.4
-0.2
0
0.2
0.4
Articles Prepositions Numbers
FtF Dyad FtF Group Dist Group * p<.05corr
elat
ion
wit
h a
gre
emen
t w
ith
gro
up
Words Expressing Positive EmotionsWords Expressing Positive Emotions
-0.5
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
Affect PosEmot PosFeel Optim
FtF Dyad FtF Group Dist Group
* p<.05corr
elat
ion
wit
h a
gre
emen
t w
ith
gro
up
Words Expressing Negative EmotionsWords Expressing Negative Emotions
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
NegEmo Anx Anger Sad
FtF Dyad FtF Group Dist Group
* p<.05corr
elat
ion
wit
h a
gre
emen
t w
ith
gro
up
Cognitive & Social Processes
-0.6-0.5-0.4-0.3-0.2-0.1
00.10.20.30.4
CogMech Insight Discrep Social Comm
Face-to-Face Dyad Face-to-Face 4 Dist-4
* p<.05
* *
corr
elat
ion
wit
h a
gre
emen
t w
ith
gro
up
Tense
-0.3-0.2-0.1
00.10.20.30.40.50.60.7
Past Present Future
FtF Dyad FtF Group Dist Group
* p<.05
*
corr
elat
ion
wit
h a
gre
emen
t w
ith
gro
up
ConclusionConclusionGroup Polarization TaskGroup Polarization Task
• No linguistic indicators for groups of four
• High performing face-to-face dyads– Avoid social words– Avoid communication words– Use future tense
ConclusionConclusionChoose TasksChoose Tasks
• Very little overlap in linguistic markers across the three Choose Tasks– Prepositions indicate poor performance for dyads
and groups doing the Desert Survival Task and dyads performing Hidden Profile Task
– Communication words indicate poor performance for dyads performing the Desert Survival and Group Polarization Tasks
– Social words indicate poor performance for dyads performing the Hidden Profile and Group Polarization Tasks
– Use of emotional words indicate poor performance for dyads performing the Desert Survival Task groups of four performing the Hidden Profile Task
• LIWC-performance relationships may be unique to each task
Predicting Group PerformancePredicting Group PerformanceTask Type 6: Resolving Conflicts of InterestTask Type 6: Resolving Conflicts of Interest
Prisoner’s DilemmaPrisoner’s Dilemma
Partner 1 Partner 2 Partner 3
Option 1
DecisionBlue $ 4.00
DecisionBlue$ 4.00
DecisionBlue$ 4.00
Option 2
DecisionGreen $ 5.00
DecisionGreen $ 5.00
DecisionGreen $ 5.00
Option 3
Decision Green $ 3.00
DecisionBlue $ 7.00
DecisionGreen $ 3.00
Option 4
DecisionGreen $ 1.00
DecisionBlue $ 5.00
DecisionBlue $ 5.00
Option 5
DecisionBlue$ 7.00
DecisionGreen $ 3.00
DecisionGreen $ 3.00
G E N E R AT E
E X E C U T E
N E G O T IAT E
C H O O S E
o f v iew p o in t
G en era tin g id e as G en era tin g p la ns
w/n o r igh t an sw
er
o f in te res t
De c id in g issu es R e so lv in g co n f lic ts R e so lv in g con f li c
ts
Re s
o lv i
n g c
o nf l
i cts
Ty p e 2 :c re a tiv ity ta s k s
Ty p e 1 :p la n n in g ta sk s
Ty p e 3 :in te lle c t iv e ta s k s
Ty p e 8 :p e rfo rm a n c e /
p s y c h o -m o to r ta s k s
Ty p e 4 :d e c is io n -m a k in g
ta s k s
Ty p e 7 :c o n te s ts /b a t tle s /
c o m p e ti tiv e ta s k s
Ty p e 6 :m ix e d -m o tiv e
ta s k s
Ty p e 5 :c o g n it iv e
c o n fl ic ts ta sk s
w/c
orre
c t a
n sw
e rs
o f p
ower
Solv
ing
p rob
lem
s Ex ecu t ing
p e r fo rmanc e tasks
Con
flic
tC
o ope
rati
on
C o n c e p tu a l B e h a v io ra l
M cG rath ’s C ircu m p lex o f G rou p Ta sk s
Ty p e 6 :m ix e d -m o tiv e
ta s k s
Pronoun Use
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
Total FirstSingular
FirstPlural
Second Third
Dist Triad FtF Triad Dist Dyad* p<.05co
rrel
atio
n w
ith
co
op
erat
ive
resp
on
ses
Cognitive Complexity
-0.4-0.3-0.2-0.1
00.10.20.30.40.50.60.7
Six-Letter Negate Inclusive Exclusive
Dist Triad FtF Triad Dist Dyad
* p<.05corr
elat
ion
wit
h c
oo
per
ativ
e re
spo
nse
s
Cognitive ComplexityCognitive ComplexityConcrete Words
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0.4
Articles Prepositions Numbers
Dist Triad FtF Triad Dist Dyad* p<.05co
rrel
atio
n w
ith
co
op
erat
ive
resp
on
ses
Words Expressing Positive Emotions
-0.3-0.2-0.1
00.10.20.30.40.50.60.7
Affect PosEmot PosFeel Optim
Dist Triad FtF Triad Dist Dyad * p<.05corr
elat
ion
wit
h c
oo
per
ativ
e re
spo
nse
s
Words Expressing Negative Emotions
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
NegEmot Anger Sad
Dist Triad FtF Triad Dist Dyad * p<.05corr
elat
ion
wit
h c
oo
per
ativ
e re
spo
nse
s
Cognitive & Social Processes
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
CogMech Insight Discrep Social Comm
Dist Triad FtF Triad Dist Dyad * p<.05
*
corr
elat
ion
wit
h c
oo
per
ativ
e re
spo
nse
s
Tense
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
Past Present Future
Dist Triad FtF Triad Dist Dyad* p<.05co
rrel
atio
n w
ith
co
op
erat
ive
resp
on
ses
ConclusionsConclusionsPrisoner’s Dilemma TaskPrisoner’s Dilemma Task
• Distributed triads that used discrepancy words were more cooperative
• Trends existed:– Cooperative distributed triads
• Avoid second person pronouns• Use future tense• Use communication words
– Cooperative face-to-face triads• Avoid second person pronouns• Use exclusive words• Express positive and negative emotions and use anger words• Avoid the future tense
– Cooperative distributed dyads• Use positive feeling words
Predicting Group PerformancePredicting Group PerformanceTask Type 8:Task Type 8: Psycho-Motor TasksPsycho-Motor Tasks
Card TaskCard Task
Make a house using as many cards as possible from several decks of playing cards
G E N E R AT E
E X E C U T E
N E G O T IAT E
C H O O S E
o f v ie w p o in t
G en era t in g id e a s G en e ra tin g p la n s
w/ n o rig h t a nsw
e r
o f in te re s t
De c id ing iss ues R eso lv in g co n f lic ts R e so lv in g con f lic
t s
Re s
o lv i
n g c
o nf l
i ct s
Ty p e 2 :c re a tiv ity t a sk s
Ty p e 1 :p la n n in g t a sk s
Ty p e 3 :in te lle c tiv e ta sk s
Ty p e 8 :p e rfo rm a n c e /
p sy c h o -m o to r ta sk s
Ty p e 4 :d e c is io n -m a k in g
ta sk s
Ty p e 7 :c o n te s ts /b a ttle s /
c o m p e titiv e ta sk s
Ty p e 6 :m ix e d -m o tiv e
ta sk s
Ty p e 5 :c o g n itiv e
c o n flic ts ta sk s
w/c
orre
c t a
n sw
e rs
o f pow
er
Solv
ing
p rob
lem
s Ex e cu tin g
p er fo r ma nc e ta s k s
Con
flic
tC
oope
r atio
n
C o n c e p tu a l B e h a v io ra l
M c G r a th ’s C irc u m p le x o f G ro u p Ta sk s
Ty p e 8 :p e rfo rm a n c e /
p sy c h o -m o to r ta sk s
Pronoun Use
-0.1
0
0.1
0.2
0.3
Total FirstSingular
FirstPlural
Second Third
* p<.05
corr
elat
ion
wit
h p
erfo
rman
ce
Cognitive Complexity
-0.4
-0.2
0
0.2
0.4
6+ Causal Negate Excl Incl
* p<.05
corr
elat
ion
wit
h p
erfo
rman
ce
Cognitive ComplexityConcrete Words
-0.4
-0.2
0
0.2
0.4
Articles Prepositions Numbers
* p<.05
corr
elat
ion
wit
h p
erfo
rman
ce
Words Expressing Emotions
-0.2
0
0.2
0.4
Affect PosFeel Anger
* p<.05
corr
elat
ion
wit
h p
erfo
rman
ce
Cognitive & Social Processes
0
0.2
0.4
0.6
CogMech Insight Discr Social Comm
* p<.05
corr
elat
ion
wit
h p
erfo
rman
ce
Tense
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0.4
Past Present Future
* p<.05
corr
elat
ion
wit
h p
erfo
rman
ce
Predicting Group PerformancePredicting Group PerformanceTask Type 8:Task Type 8: Psycho-Motor TasksPsycho-Motor Tasks
Radio Assembly TaskRadio Assembly Task
As a group, we would like you to assemble the AM portion of a radio using a Radio Kit from Radio Shack. To assemble the AM portion of the radio, you will need to insert dozens of components into different places on the circuit board and then connect each component to the others in the proper manner.
fifteen minutes
G E N E R AT E
E X E C U T E
N E G O T IAT E
C H O O S E
o f v ie w p o in t
G en era t in g id e a s G en e ra tin g p la n s
w/ n o rig h t a nsw
e r
o f in te re s t
De c id ing iss ues R eso lv in g co n f lic ts R e so lv in g con f lic
t s
Re s
o lv i
n g c
o nf l
i ct s
Ty p e 2 :c re a tiv ity t a sk s
Ty p e 1 :p la n n in g t a sk s
Ty p e 3 :in te lle c tiv e ta sk s
Ty p e 8 :p e rfo rm a n c e /
p sy c h o -m o to r ta sk s
Ty p e 4 :d e c is io n -m a k in g
ta sk s
Ty p e 7 :c o n te s ts /b a ttle s /
c o m p e titiv e ta sk s
Ty p e 6 :m ix e d -m o tiv e
ta sk s
Ty p e 5 :c o g n itiv e
c o n flic ts ta sk s
w/c
orre
c t a
n sw
e rs
o f pow
er
Solv
ing
p rob
lem
s Ex e cu tin g
p er fo r ma nc e ta s k s
Con
flic
tC
oope
r atio
n
C o n c e p tu a l B e h a v io ra l
M c G r a th ’s C irc u m p le x o f G ro u p Ta sk s
Ty p e 8 :p e rfo rm a n c e /
p sy c h o -m o to r ta sk s
Pronoun Use
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
Total FirstSingular
FirstPlural
Second Third
Dyad Triad* p<.05
corr
elat
ion
wit
h t
ime
to c
om
ple
te r
adio
Cognitive Complexity
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
6+ Causal Negate Excl Incl
Dyad Triad
* p<.05
corr
elat
ion
wit
h t
ime
to c
om
ple
te r
adio
*
Cognitive ComplexityConcrete Words
-0.8-0.6-0.4-0.2
00.20.40.60.8
1
Articles Prepositions Numbers
Dyad Triad* p<.05
corr
elat
ion
wit
h t
ime
to c
om
ple
te r
adio
*
Words Expressing Positive Emotions
-0.8
-0.6
-0.4
-0.2
-0
0.2
0.4
0.6
Affect PosEmot PosFeel
Dyad Triad
* p<.05corr
elat
ion
wit
h t
ime
to c
om
ple
te r
adio
Words Expressing Negative Emotions
-0.8-0.6-0.4-0.2
-00.20.40.60.8
1
NegEmot Anger Sad
Dyad Triad
* p<.05
corr
elat
ion
wit
h t
ime
to c
om
ple
te r
adio
Cognitive & Social Processes
-0.8-0.6-0.4-0.2
00.20.40.60.8
1
CogMech Insight Discrep Social Comm
Dyad Triad
* p<.05
corr
elat
ion
wit
h t
ime
to c
om
ple
te r
adio
Tense
-1-0.8-0.6-0.4-0.2
00.20.40.60.8
Past Present Future
Dyad Triad
* p<.05
corr
elat
ion
wit
h t
ime
to c
om
ple
te r
adio
ConclusionConclusionRadio Assembly TaskRadio Assembly Task
• High Performing Dyads– Avoid Prepositions– Trend to:
• Use pronouns (esp. first person singular and second person)
• Use negations• Avoid articles• Use positive feeling
words• Avoid anger words• Use words expressing
cognitive and social processes
• Use present tense
• High Performing Triads– Use words with six or
more letters• Trend to:
– Avoid causal words– Avoid exclusive words– Use prepositions– Use numbers– Avoid words expressing
cognitive mechanisms– Avoid discrepancy words– Avoid the present and
future tenses
General ConclusionsGeneral Conclusions
• LIWC variables have been found to be useful in predicting group performance BUT the linguistic categories that predict performance differ by task, task type, group size, and communication medium
• Relationship to group processes have yet to be discovered
Future ResearchFuture Research
• Determine LIWC variables which can indicate group relations (e.g., trust, use of first person singular pronouns to predict status)
• Determine usefulness of LIWC in assessing group readiness
• Examine usefulness of LIWC in determining level of group development