Cases_(Chapter_13_pp.344-347)

download Cases_(Chapter_13_pp.344-347)

of 4

Transcript of Cases_(Chapter_13_pp.344-347)

  • 8/18/2019 Cases_(Chapter_13_pp.344-347)

    1/4

      44

    Part

    Sample Design for

    Data

    CoUection

    and

    Sample Size

    C A S E S

    CASE

    4  1

    Riverside County Humane Society B)

    Like a lot of nonprofits, Riverside County Humane So

    ciety RC HS)

    is

    being expected to do more (services)

    with less (funding and donations). In an effort to halt

    further decline in the quality of its services and to pro

    vide better care for the pets at the center, the Mem

    bership Committee of the board of directors began

    making plans for a member/contributor drive. The or

    ganized drive was the first of its kind for the local

    chapter and the committee members wanted it to be as

    productive as possible.

    As the plans began to

    evolve,

    the committee realized

    that the organization had only scattered bits and pieces

    of information about its current members. It did have a

    list of members and contributors for the last five years

    that had been compiled by the RCHS staff. In addition, it

    had access to the results of a survey that had been done

    by a staff member several years earlier that focused on

    member usage of shelter facilities

    and

    their opinions of

    shelter services and programs. However, the organiza

    tion had only sparse knowledge

    of

    the profile

    of

    its

    typical member

    and

    contributor, why they belonged or

    contributed, how long they had been associated with the

    humane society, how the services of the humane society

    could

    be

    improved, and so

    on.

    The committee members

    believed information on these issues was important to

    the conduct of a successful membership drive, and thus

    they commissioned research to secure t

    Initial contacts with other humane society chapters

    and interviews with some RCHS staff and board

    members produced a number of hypotheses regarding

    who is likely to become a member or contributor, why,

    how much people are likely to give, and so

    on.

    The

    researchers are interested

    in

    examining these hy

    potheses through a mail survey sent to current mem

    bers and contributors. (See Case 2.2, Riverside County

    Humane Society (A) for details.)

    Sampling P lan

    For the last five years, the RCHS had maintained a

    master list of members

    and

    contributors. Contributors

    were those who had sent a donation to RCHS but had

    not filled out an official form making them members,

    which essentially entitled them to receive RCHS's

    newsletter. The separate list of members contained all

    those who had expressed interest in membership and

    who were receiving the newsletter. Both lists were

    alphabetical. The contributor list included the amount re

    ceived from each person or business, but not the number

    of times the person or business

    had

    given during the last

    five years. The member list showed the number of years

    each organization or person had belonged.

    For purposes of the study, all names of businesses or

    other organizations were deleted

    and

    a separate sample

    was taken from each list Approximately 1,050 people

    were on the member list and 300 on the contributor list

    The researchers decided to take 120 names from the

    member list

    and

    50 from the contributor list. They iden

    tified those to be sent questionnaires by drawing two

    random numbers-3 and

    5 using

    a random number

    table. They then sent questionnaires to the 3 '

    d

    11th,

    19th,

    and so on, person on the member l s ~ and the 5th,

    11th,

    1

    ?1h

    and so on, person on the contributor list

    QU STIONS

    1. What is the sampling frame and is it a good frame

    for the target population?

    2. What type of sample is being used?

    3. Can you think of some ways in which the sample

    could

    be

    improved?

    CASE

    2

    Sampling Lead Use

    rs

    for New

    Products

    The new, hot thing

    in

    testing new product concepts

    and

    product specifications 1s to bring in samples of lead

    users

    .

    the customers who are ··innovators ' in terms

    of diffusion of innovation, or those who precede the

    early adopters·'-they're the earliest of adopters. The

    idea is that these customers are so highly involved in

    this product category that they're likely to encounter

    needs and desire benefits before the rest of the mar

    ketplace might even recognize that there's an issue or

    an unmet need. Their customer involvement is also so

    high that they're very likely to agree to be involved in

    beta testing products or giving their expert opinions

    about product design and the like.

    QU STIONS

    1.

    What kind of information can you obtain from a

    ··lead-user

    .

    sample?

    This document is authorized for use by

    Raymond a

    , from 3/14/2016 to 4/20/2016, in the course:

    MGMT 686.001: Marketing Research - Porter Spring 2016), Rice University.

    Any unauthorized use or reproduction of this document is strictly prohibited.

  • 8/18/2019 Cases_(Chapter_13_pp.344-347)

    2/4

     2

    . Is this a probability sample? A cluster sample? A

    stratified sample? A convenience sample?

    3. What limitations might there be to a lead-user

    sample?

    CASE4 .3

    Sampling

    from

    Your

    Loyalty

    Database

    Cain is a holding company that publishes magazines. It

    has a large subscription database, across all its maga

    zines. It wants to use th is database to get feedback on

    some ideas about a new magazine, a new loyalty pro

    gram, and a new pricing policy as it makes its period

    icals avai lab le online.

    Here's what they're thinking

    of

    doing. They have the

    entire database-that's their sampling frame. They're

    going to split the customers into good ones (sub

    scribe to

    two

    or more magazines), current customers'

    (subscribe to one), and ·'lapsed customers·· those who

    subscribed to one or more, more than six months ago

    but did not keep up their subscription).

    Then, the database is huge and they only want a

    sample of 300 of each group good/ currenV lapsed

    customers). So once the database is parsed into these

    t

    hr

    ee groups, they're going to take every 50

    1

    h

    customer

    in the database until they have 300, then stop and go

    onto the next group.

    QU STIONS

    1. What kind of sampling plan is this?

    2. Evaluate the sampling plan.

    Table

    4 .4 . 1

    The Estimates for

    the

    U.S. Samples

    Cases

    CASE4 .4

    International Differences in

    the

    Cost

    of

    Data

    345

    A multinational bank, MNB, was interested in measur

    ing customer satisfaction with its consumer banking

    services and financial products. The managers at

    MNB

    Corporate differed in their opinions regarding

    the

    form

    of the optimal customer satisfaction study: survey,

    focus group, interview, and

    so on.

    So

    as

    a starting point,

    MNB

    commissioned bids from marketing research

    firms

    in

    the United States to describe how they would

    approach studying their

    U.S.

    banking customers. In

    particular, they asked the bidding firms to

    offer

    cost

    estimates, that

    is,

    how many customers could be sam

    pled given the proposed budget

    of 1

    5,

    000

    for th is

    re

    search project using different techniques.

    The bids covered a variety

    of

    research methods,

    which differed in their costs.Th

    ey

    con

    si

    dered a personal

    interview method in which the customer would be i

    n-

    tercepted in the bank

    an

    d asked several questions re

    garding service and satisfaction. They compared that

    method to the costs of sending out mail surveys to

    current customers. Finally, they explored the efficiency

    of

    placing small, postcard-sized surveys at each teller

    station that the

    ba

    n

    ki

    ng customer might pick up and

    complete and return at their leisure.

    MNB gathered the marketing research firms' propo

    sals to begin to make a decision

    of

    how the bank should

    approach its customers. In terms

    of

    outlay expenses,

    more postcards could be printed less expensively than

    surveys sent or interviewers staffing each of the local

    ban

    k branches. However, in terms of

    re

    sponse rates,

    somewhat fewer people turned down the personal in

    terview than who returned the mail survey or postcard

    survey. Table 4.4.1 presents the comparative estimates

    Interview

    Mail

    Survel:'.

    Postcards

    a.

    Research bu dget

    15,000

    15 ,000 15,0

    00

    b. Cost per

    co

    n

    tact

    25

    2

    .50

    S.25

    c. Prospects reached•

    2,400 6,000

    60

    ,000

    d . R

    espo

    nse rate

    5% 3%

    1.5%

    e. Estimated net s

    am

      e size (c x d ) 120 180

    90

    0

    f. Effective cost per capita (a/ e) 1 25.00 83.33

    16.6

    7

    ·F

    or

    mail and postcard, c

    =

    alb; for interview

    s,

    cost

    is 25

    per hour, times a f

    1ve·

    hour day,

    = 1

    25

    per d

    ay,

    for each

    in

    terviewer. Each of

    12

    1nterv1ewers (spread across the area local bank branches) wou ld spend

    two

    weeks ( 10 bank days) approaching and in terviewing customer

    s.

    Each interviewer on each day would target 20 bank customers, on average.

    his document is authorized for use byRaymond a, from 3/14/2016 to 4/20/2016, in the course:

    MGMT 686.001: Marketing Research - Porter Spring 2016), Rice University.

    Any unauthorized use or reproduction of this document is strictly prohibited.

  • 8/18/2019 Cases_(Chapter_13_pp.344-347)

    3/4

    346

    Part4

    Sample Design for Data Collection and Sample Size

    J

    - : . . ~ ~ ; : : ~ .

    :.,

    , • ;,_ • - •• , . . .. , • . . • : -

    Tabl

    e

    4  4  2 The

    Pr.ojections for Indonesia

    Interview M  l Surv

    er

    P

    os

    tcards

    a. Research budget 15 ,000 15 ,000 15 ,000

    b. Cose per concact

    12*

    2.5 0 .25

    c.

    Prospects reached

    5,000

    6,000 60 ,000

    d. Response rate

    20%* 1.5%* .5%

    e.

    Estimated

    nee

    sample size (c x d)

    1,000 90 300

    f. Effective cost per capita (a/e)

    15 .

    00

    166.67

    50.00

    'These estimates differ from those for conducting the research n the U.S . labor costs for mterv1

    ew1ng

    are much less, propens  t

    1es

    for customers to acquiesce and be interviewed are much greater mail surveys are somewhat less efficient because CD-ROM databases on

    addresses are less accurate and postcards are an unfam1l1ar format and are therefore rarely filled out. Given the changes n the interviewing

    parameters, If the project duration

    is

    still two weeks, 25 1nterv1ewers can be deployed, so

    5 000

    customers would

    be

    approached.

    for the three different techniques. The table shows

    cl

    early that

    in

    the

    U.S.

    , the postcard technique appears

    to be the most cos t effect

    ive.

    Thus the bank managers

    are considering implemen ting this research tool.

    MNB's second conce

    rn is

    with a few of its satellite

    locations-in particular, the news from abroad

    is

    that

    the Indonesian banking customers are not happy-and

    MNB w

    an

    ts to underst

    an

    d what is going

    on.

    Its first

    assumpt

    io

    n w

    as

    that the bank should proceed with the

    pos t

    ca

    rd me thodology to be able to compare the re

    sults in Indonesia to those from the States. However,

    they conducted

    som

    e preliminary investigations

    and

    found the costs of the methods to

    be

    quite different.

    Th

    e bank plans to proceed with personal

    in

    terviews

    in

    Indonesia. Co nsider Tables

    4.4.1 and

    4

    .4.2

    and answer

    the questions that foll

    ow.

    QU STIONS

    1. Under what co

    nd

    itions might it matter that one

    method is used in one country and another method

    is used in

    an

    other?

    2. What is the target population under investigation

    both in the U.S. and in Indonesia? What are the

    dif ferences between the sampling frames of each

    of

    the three techniques? What customers will each

    techniqu

    e miss?

    3. What other issues must the bank managers con

    sider in addition to the cost efficiencies of the three

    methods?

    C S

     

    .5

    nline Sampl

    es

    You're working for a lar

    ge

    pharmaceutical company on

    a brand management team,

    and

    your team

    is

    responsible for building two initiatives. Fi rst you'

    re

    creating a Web site for social support-blogging

    by

    users and carriers to talk about their drug experiences

    but also how their disease impacts their lives, how they

    cope, things they've learned, just general, "we'

    re

    here

    for you'· kind of suppor

    t.

    T

    he

    second initiative is based

    on several pharma-firms' recent successes with flavors

    for children's medici

    nes. You

    r drugs are for adults but

    you

    figur

    e

    if it works for kids, why not for adu l

    ts

    too,

    so

    you're exploring the possible desire for several flavors

    (cherry, strawberry, pina colada, etc.).

    Your team would like to get some good consumer

    data before proceeding, because each initiative would

    eventually require a good level

    of

    financial backing.

    You've got a beta version of the Web site developed,

    you're ready for potential-user feedback,

    and

    your

    chemists have created several flavors that are

    palatable, which your t

    eam

    is ready to test.

    Your team meets with your boss and his boss. Your

    team has proposed a combination of marketing re

    search

    me

    thods including a few focus groups, some

    on line surveys,

    and

    some mall-intercepted interviews.

    Your boss is supportive, as you all knew he would be

    from previous meetings, but he's keeping quiet be

    cause this meeting

    is

    the first during which the ideas

    are being presented to the big boss.

    The big boss likes the focus group idea, doesn't

    really like the mall-intercept idea but mostly wants re

    sults quickly ("yesterday " ) and cheaply (ideally

    free-

    no line

    on

    the marketi

    ng

    budget). You all know he's

    exaggerating, b

    ut

    he's like that, and you know he's not

    exaggerating by much. So , he's pushing doing the

    testing of the Web site and the flavors via your on line

    panel of past customers who have redeemed pre

    scriptions for your drugs.

    This document is authorized for use byRaymond a, from 3/14/2016 to 4/20/2016, in the course:

    MGMT 686.001: Marketing Research - Porter Spring 2016), Rice University.

    Any unauthorized use or reproduction of this document is strictly prohibited.

  • 8/18/2019 Cases_(Chapter_13_pp.344-347)

    4/4

      E

    STIONS

    1. The

    onl in

    e panel was, of course, part of your pro

    posal. Shall you just back down  and do what the

    big boss wants?

    2. What are the characteristics of the e-panel likely to

    be?

    3.

    What kind of information can you obtain from the

    online survey?

    4.

    Will

    yo

    u

    be

    ready to launch both initiatives after

    conducting solely the online survey?

    SE

    4  6

    Sampling

    ambl

     

    ng

    Americans spend nearly 50 billion annually on gam

    bling. That expenditure is more than three times t

    he

    amount spent on going to the movies and theme parks,

    as

    alternative means

    of

    entertainment. State-run lot

    teries

    and

    casino games are the most popular legal

    games of chance.

    Ga

    mbling traditionally held negative connotations,

    being associated with immoral or even criminal behavior.

    Cri t ics worry that legalized gambling can encourage

    compul

    sive

    gambl

    ers,

    that it may encourage people to

    gamble who can least afford to do

    so,

    and that casinos

    bring an undesirable element to the surrounding

    neighbor

    hood.

    Legalized gambling, however, especially in the form

    of state lotteries,

    has

    largely sanitized the image of

    gaming

    be

    h

    av

    i

    or.

    Among people who abstain from

    gamblin

    g,

    fewer peo

    pl

    e cite moral or relig ious objec

    tions, instead offering practical reasons: e.g. they don't

    want to spend the money, or they don't have the money

    to spend. Proponents claim casinos create jobs and

    provide revenue for education that would otherwise be

    raised

    by

    tax hikes.

    Secondary data suggest that there

    is no

    particular

    demographic profile of a gambler-people of all walks

    of life e.g. age and income) enjoy casinos and lotteries.

    Beyond demographics, the commonly held motivations

    appear to be a desire to win a large amount of money,

    and

    a quick and relatively inexpen

    sive

    form

    of

    entertainment.

    The heterogeneity of the demographics of gamblers

    and the homogeneity of their motives have left some

    casino managers perplexed

    as

    to how t

    he

    consumer

    Ca

    ses 47

    market might be segmented. One hypothesis, based on

    collective wisdom,

    is

    that novice gamblers tend to pre

    fer slot machines because they are simple, whereas

    more experienced gamblers prefer games like black

    jack, baccarat, and craps because they are more stra

    tegic in nature. Another frequent assumption is that

    people who buy lottery tickets are different from people

    who go to casinos,

    and

    that the two types of games

    satisfy different needs.

    The Internet introduces yet another medium in

    which a consumer might gamble. Although the number

    of gamblers online are far fewer than those who

    frequent casinos or lotteries, and online gambling rev

    enues are far less than nonelectronic games (approxi

    mately only

    2

    billion), the online gambling industry is

    expected to enjoy rap id growth.

    If a gaming industry representati

    ve

    were to come to

    you and say , I'd like to do some interviews. I want to

    know more about what kinds of games my casino vi

    si

    tors want me to provide. I want to know how much floor

    space to allot

    to

    slots versus blackjack tables. I want to

    know why they come to any casino, or my casino, rather

    than to the movies or something. Maybe I should just

    sell lottery t ickets. How wo

    ul

    d you address the fol

    lowing issues?

    QU STIONS

    1. What is the relevant population?

    2. What would you

    re

    commend in terms of a sampling

    plan- would

    you intercept people in (or entering or

    exi ting) casinos? Would you talk

    to

    peo

    pl

    e buying

    lottery tickets at convenience stores? Would you

    interview people stroll i

    ng

    along

    in

    a shopping mall?

    What sampli

    ng

    frame does each of these locations

    presume?

    3.

    If you wanted to verify the aforementioned as

    sumptions (for example, that demographics do not

    matter, or that novice versus experienced gamblers

    have different game preferences), how would you

    modify your sampling plan?

    4. How would you try to assess objectivel y whether

    there are moral or

    re

    ligious concerns against gam

    bl

    ing? What kinds of sampling plans would you

    need to avoid, given the likel ihood that they would

    shape your conclusions pro or con?

    This document is authorized for use by Raymond a, from 3/14/2016 to 4/20/2016, in the course:

    MGMT 686.001: Marketing Research - Porter Spring 2016), Rice University.

    Any unauthorized use or reproduction of this document is strictly prohibited.