Case One: Attiring Situation1

Case 1

Attiring Situation

Objectives: This case allows the student to create hypotheses and conduct statistical analyses to test them using data from an experiment.

Summary: RESERV is a national placement firm specializing in putting retailers and service providers together with potentialemployees who fill positions at all levels of the organization—from entry-level positions to senior management positions. One specialty clothing store chain has adopted a very flexible dress code and is interested in examining if the appearanceof potential employees influences customers. The retailer also is interested in customer integrity. The senior research associate conducts an experiment to examine relevant research questions including:

RQ1:How does employee appearance affect customer purchasing behavior?

RQ2:How does employee appearance affect customer ethics?

A laboratory experiment is designed in which two variables are manipulated in a between-subjects design: employee attire (professional/unprofessional) and manner with which the employee tries to gain extra sales (soft close/hard close). Subjects’ biological sex was recorded and included as a blocking variable. Four dependent variables are included: TIME (0-10 minutes), SPEND ($0-$25), and KEEP ($0-$25). Additionally, several variables were collected following the experiment that tried to capture how the subject felt during the exercise. All of these items were gathered using a 7-item semantic differential scale.

The experiment was conducted in a university union, and subjects were recruited to participate as customers who had just purchased some dress slacks and a shirt in a mock retail environment. The employee was to complete the transaction and try to sell the customer some of several accessory items displayed at the counter. Each subject was randomly assigned to one of four conditions where the employee was either:

Dressed professionally and used a soft close.
Dressed unprofessionally and used a soft close.
Dress professionally and used a hard close.
Dress unprofessionally and used a hard close.

The researcher wishes to use this information to explain how employee appearance encourages shoppers to continue shopping (TIME) and spend money (SPEND). Each subject was given $25 (in one-dollar bills) which they were allowed to spend on accessories. Subjects were not told what to do with any of the money they did not spend, so the other dependent variable, KEEP, measured how much of the money a subject kept after returning the questionnaire.

Questions

1.Develop at least three hypotheses that correspond to the research questions.

Students’ hypotheses will vary. However, some possible hypotheses are:

H1:Customers will spend less if the employee is dressed professionally.

H2:Customers will spend more if the employee uses a hard close.

H3:Customers will keep less of the money if the employee is dressed professionally.

Students might hypothesize that there will be differences between males and females. For example:

H4:Males will spend less than females if the employee is dressed professionally.

Ideally, hypotheses would be developed based on theory or a proposed model that would logically lead to a specific hypothesis.

2.Test the hypotheses using an appropriate statistical approach.

The appropriate statistical approach will depend on the hypotheses students develop. ANOVA is the appropriate statistical approach for the hypotheses given above.

To test H1 above, ANOVA is appropriate:

Group Statistics
X1 / N / Mean / Std. Deviation / Std. Error Mean
SPEND / PROF_ATTIRE / 50 / 3.60 / 3.110 / .440
UNPROF_ATTIRE / 50 / 14.00 / 7.371 / 1.042
Independent Samples Test
Levene's Test for Equality of Variances / t-test for Equality of Means
95% Confidence Interval of the Difference
F / Sig. / t / df / Sig. (2-tailed) / Mean Difference / Std. Error Difference / Lower / Upper
SPEND / Equal variances assumed / 29.103 / .000 / -9.192 / 98 / .000 / -10.400 / 1.131 / -12.645 / -8.155
Equal variances not assumed / -9.192 / 65.914 / .000 / -10.400 / 1.131 / -12.659 / -8.141

These results suggest that customers spend less when the employee is dressed professionally, providing support for H1. However, there is no significant different on the amount spend due to the type of close used by the employee (H2):

Group Statistics
X2 / N / Mean / Std. Deviation / Std. Error Mean
SPEND / SOFT_CLOSE / 50 / 8.28 / 8.064 / 1.140
HARD_CLOSE / 50 / 9.32 / 7.322 / 1.035
Independent Samples Test
Levene's Test for Equality of Variances / t-test for Equality of Means
95% Confidence Interval of the Difference
F / Sig. / t / df / Sig. (2-tailed) / Mean Difference / Std. Error Difference / Lower / Upper
SPEND / Equal variances assumed / 1.208 / .274 / -.675 / 98 / .501 / -1.040 / 1.540 / -4.097 / 2.017
Equal variances not assumed / -.675 / 97.101 / .501 / -1.040 / 1.540 / -4.097 / 2.017

Similarly, the results do not support H3, which hypothesized that customers would keep less money if the employee was dressed professionally. There was no significant difference on how much money customers kept due to the employee’s attire:

Group Statistics
X1 / N / Mean / Std. Deviation / Std. Error Mean
KEEP / PROF_ATTIRE / 50 / 4.40 / 3.110 / .440
UNPROF_ATTIRE / 50 / 6.62 / 3.979 / .563
Independent Samples Test
Levene's Test for Equality of Variances / t-test for Equality of Means
95% Confidence Interval of the Difference
F / Sig. / t / df / Sig. (2-tailed) / Mean Difference / Std. Error Difference / Lower / Upper
KEEP / Equal variances assumed / 2.735 / .101 / -3.108 / 98 / .002 / -2.220 / .714 / -3.637 / -.803
Equal variances not assumed / -3.108 / 92.601 / .002 / -2.220 / .714 / -3.638 / -.802

Finally, H4 stated that males would spend less than females if the employee was dressed professionally, which is supported by the results given below:

Between-Subjects Factors
Value Label / N
X1 / 0 / PROF_ATTIRE / 50
1 / UNPROF_ATTIRE / 50
Gender / 0 / MALE / 41
1 / FEMALE / 59
Tests of Between-Subjects Effects
Dependent Variable:SPEND
Source / Type III Sum of Squares / df / Mean Square / F / Sig.
Corrected Model / 2988.385a / 3 / 996.128 / 33.535 / .000
Intercept / 6750.794 / 1 / 6750.794 / 227.266 / .000
X1 / 2112.722 / 1 / 2112.722 / 71.125 / .000
Gender / 25.206 / 1 / 25.206 / .849 / .359
X1 * Gender / 220.312 / 1 / 220.312 / 7.417 / .008
Error / 2851.615 / 96 / 29.704
Total / 13584.000 / 100
Corrected Total / 5840.000 / 99
a. R Squared = .512 (Adjusted R Squared = .496)
EXP1 * Gender
Dependent Variable:SPEND
EXP1 / Gender / Mean / Std. Error / 95% Confidence Interval
Lower Bound / Upper Bound
PROF_ATTIRE / MALE / 1.871 / .979 / -.072 / 3.814
FEMALE / 6.421 / 1.250 / 3.939 / 8.903
UNPROF_ATTIRE / MALE / 15.800 / 1.723 / 12.379 / 19.221
FEMALE / 13.550 / .862 / 11.839 / 15.261

3.Suppose the researcher is curious about how the feelings captured with the semantic differentials influence the dependent variables SPEND and KEEP. Conduct an analysis to explore this possibility. Are any problems present in testing this?

The eight semantic differential variables were regressed on each dependent variable: SPEND and KEEP, and results are given on the following pages. There is a problem with multicollinearity (note the high VIFs), so factor analysis would be useful.
Dependent variable = SPEND, model is not significant (F = 0.62, p = 0.759):

Variables Entered/Removed
Model / Variables Entered / Variables Removed / Method
1 / SD8, SD2, SD1, SD5, SD3, SD7, SD4, SD6a / . / Enter
a. All requested variables entered.
Model Summary
Model / R / R Square / Adjusted R Square / Std. Error of the Estimate
1 / .231a / .053 / -.033 / 7.805
a. Predictors: (Constant), SD8, SD2, SD1, SD5, SD3, SD7, SD4, SD6
ANOVAb
Model / Sum of Squares / df / Mean Square / F / Sig.
1 / Regression / 302.072 / 8 / 37.759 / .620 / .759a
Residual / 5360.958 / 88 / 60.920
Total / 5663.030 / 96
a. Predictors: (Constant), SD8, SD2, SD1, SD5, SD3, SD7, SD4, SD6
b. Dependent Variable: SPEND
Coefficientsa
Model / Unstandardized Coefficients / Standardized Coefficients / t / Sig. / Collinearity Statistics
B / Std. Error / Beta / Tolerance / VIF
1 / (Constant) / 39.430 / 15.430 / 2.555 / .012
SD1 / .441 / .436 / .117 / 1.012 / .314 / .809 / 1.236
SD2 / -.067 / .718 / -.018 / -.094 / .926 / .289 / 3.456
SD3 / -1.184 / 1.030 / -.311 / -1.150 / .253 / .147 / 6.789
SD4 / -2.570 / 1.772 / -.647 / -1.450 / .150 / .054 / 18.496
SD5 / -.771 / 1.100 / -.159 / -.701 / .485 / .209 / 4.778
SD6 / -3.325 / 1.927 / -.837 / -1.726 / .088 / .046 / 21.871
SD7 / .124 / 1.856 / .021 / .067 / .947 / .109 / 9.147
SD8 / -.283 / 1.161 / -.052 / -.244 / .808 / .237 / 4.216
a. Dependent Variable: SPEND
Collinearity Diagnosticsa
Model / Dimension / Eigenvalue / Condition Index / Variance Proportions
(Constant) / SD1 / SD2 / SD3 / SD4 / SD5 / SD6 / SD7 / SD8
1 / 1 / 7.750 / 1.000 / .00 / .00 / .00 / .00 / .00 / .00 / .00 / .00 / .00
2 / .899 / 2.936 / .00 / .00 / .01 / .01 / .00 / .00 / .00 / .00 / .00
3 / .165 / 6.857 / .00 / .84 / .01 / .00 / .00 / .00 / .00 / .00 / .01
4 / .084 / 9.612 / .00 / .01 / .02 / .00 / .00 / .02 / .01 / .01 / .18
5 / .055 / 11.825 / .00 / .01 / .72 / .21 / .00 / .00 / .00 / .00 / .00
6 / .022 / 18.767 / .01 / .01 / .07 / .27 / .11 / .44 / .04 / .00 / .00
7 / .017 / 21.616 / .00 / .08 / .11 / .33 / .22 / .51 / .05 / .02 / .00
8 / .007 / 33.596 / .00 / .00 / .03 / .01 / .02 / .01 / .12 / .93 / .77
9 / .002 / 64.973 / .99 / .04 / .02 / .16 / .64 / .01 / .77 / .04 / .04
a. Dependent Variable: SPEND

Dependent variable = KEEP, also not significant (F = 0.388, p = 0.924):

Variables Entered/Removed
Model / Variables Entered / Variables Removed / Method
1 / SD8, SD2, SD1, SD5, SD3, SD7, SD4, SD6a / . / Enter
a. All requested variables entered.
Model Summary
Model / R / R Square / Adjusted R Square / Std. Error of the Estimate
1 / .185a / .034 / -.054 / 3.823
a. Predictors: (Constant), SD8, SD2, SD1, SD5, SD3, SD7, SD4, SD6
ANOVAb
Model / Sum of Squares / df / Mean Square / F / Sig.
1 / Regression / 45.408 / 8 / 5.676 / .388 / .924a
Residual / 1285.977 / 88 / 14.613
Total / 1331.384 / 96
a. Predictors: (Constant), SD8, SD2, SD1, SD5, SD3, SD7, SD4, SD6
b. Dependent Variable: KEEP
Coefficientsa
Model / Unstandardized Coefficients / Standardized Coefficients / t / Sig. / Collinearity Statistics
B / Std. Error / Beta / Tolerance / VIF
1 / (Constant) / 10.477 / 7.557 / 1.386 / .169
SD1 / .072 / .213 / .039 / .336 / .737 / .809 / 1.236
SD2 / .193 / .352 / .107 / .549 / .584 / .289 / 3.456
SD3 / .007 / .504 / .004 / .014 / .989 / .147 / 6.789
SD4 / -.965 / .868 / -.501 / -1.112 / .269 / .054 / 18.496
SD5 / -.402 / .539 / -.171 / -.746 / .458 / .209 / 4.778
SD6 / -.505 / .944 / -.262 / -.535 / .594 / .046 / 21.871
SD7 / .473 / .909 / .165 / .521 / .604 / .109 / 9.147
SD8 / -.113 / .569 / -.043 / -.200 / .842 / .237 / 4.216
a. Dependent Variable: KEEP
Collinearity Diagnosticsa
Model / Dimension / Eigenvalue / Condition Index / Variance Proportions
(Constant) / SD1 / SD2 / SD3 / SD4 / SD5 / SD6 / SD7 / SD8
1 / 1 / 7.750 / 1.000 / .00 / .00 / .00 / .00 / .00 / .00 / .00 / .00 / .00
2 / .899 / 2.936 / .00 / .00 / .01 / .01 / .00 / .00 / .00 / .00 / .00
3 / .165 / 6.857 / .00 / .84 / .01 / .00 / .00 / .00 / .00 / .00 / .01
4 / .084 / 9.612 / .00 / .01 / .02 / .00 / .00 / .02 / .01 / .01 / .18
5 / .055 / 11.825 / .00 / .01 / .72 / .21 / .00 / .00 / .00 / .00 / .00
6 / .022 / 18.767 / .01 / .01 / .07 / .27 / .11 / .44 / .04 / .00 / .00
7 / .017 / 21.616 / .00 / .08 / .11 / .33 / .22 / .51 / .05 / .02 / .00
8 / .007 / 33.596 / .00 / .00 / .03 / .01 / .02 / .01 / .12 / .93 / .77
9 / .002 / 64.973 / .99 / .04 / .02 / .16 / .64 / .01 / .77 / .04 / .04
a. Dependent Variable: KEEP

4.Critique the experiment from an internal and external validity viewpoint.

The approach used a laboratory experiment, which allows the researcher more complete control over the research setting and extraneous variables. Thus, the experiment has high internal validity, which was defined in chapter 11 as the extent that an experimental variable is truly responsible for any variance in the dependent variable. However, there is a tradeoff with respect to external validity, which is the accuracy with which experimental results can be generalized beyond the experimental subjects. Ideally, results from lab experiments would be followed up with some type of field test. Some students also may mention that the experiment used student surrogates, which might diminish external validity. However, the retail client permitted causal attire with the idea that younger customers could better identify with store employees, most of whom are younger than average.

5.What conclusions would be justified by management regarding their employee appearance policy?

The results suggest that this retailer should continue the flexible dress code as unprofessional attire resulted in higher expenditures than when the employee was dressed professionally. However, management must be careful drawing conclusions from this study due to the artificial nature of the experiment and the sample used.