Harvard Business School Cases for APPL 642

 Instructions for SPSS Data analysis case reports


Each student will analyze the SPSS data set provided for each case and submit to the instructor, a report of the results and findings on the due date indicated in the syllabus schedule.

Some important notes:

 On first page:  

The report must:


1. Conform to APA style including tables and/or figures that are incorporated into the document (in the body of the text, not at the end)

    Note: use SPSS "academic style (in Table Looks option)

2. Be submitted online to Sakai by the due date


The report MUST contain the following sections:

Note: 1: Audience: Report should be directed to managers who are expected to have a genneral understanding of descriptive statistics. However, you should explain in plain English the more arcane concepts of alpha (Cronbach), ANOVA,  t-test, Chi squared, etc.  Also explain your "approaches, strategies, tools," etc.,  such as JCM, VIE. but DO NOT refer to them them as "THEORY". Guess why!

Note 2: Use past tense since this a report to management for what has been done. (future tense would be used for a proposal or RFP)




Objectives. The purpose of conducting an annual SEP Organizational survey was to assess employee attitudes toward SEP management policies and change initiatives, and satisfaction with important aspects of their jobs. The first annual survey was intended to establish a baseline of opinion for comparison with subsequent administrations of the survey. This will enable SEP management to assess the effectiveness of organizational change initiatives. It also provides valuable information to determine where improvements can be made to increase organizational morale, effectiveness and efficiency of operations.

Several specific attitudes were targeted for assessment, particularly with regard to how employees felt about the pay, promotional opportunities, benefits, supervision, contingent rewards, job conditions, coworkers, the nature of the work itself, and effectiveness of communication. By using a standardized job satisfaction inventory in the survey, management is able to compare SEP worker opinions with other companies and as well, internal comparisons among jobs within SEP.

A sample of SEP employees were administered the Specter Job Satisfaction Survey. Additional information on the Specter Job Satisfaction survey can be found on-line.

  • Paul Spector's Job Satisfaction Scale Info

    Non-exempt workers (Line workers), made up approximately 80% of the sample. Managers (Business managers and Line supervisors) comprised about 20 % including Top Managers (Director level staff) .



    The results of the survey have been entered into an SPSS file named Sedalia A Team.sav (...B Team.sav, etc.). The variables in the data file include: The 36 individual item responses on the Job Satisfaction Survey and the employee's job class (job2). Note that the items that need to be reversed scored have already been been recoded and reordered by facets (in groups of 4). All you will need to do is sum the four items for each of the 9 facets and the Total score using the SPSS compute statement.


    For the data output to use in your report include:

    1. Frequencies and percentages for job2.

    2. Establish psychometric properties: Inter-correlations among facets to see if the facets are relatively independent.

    3. Establish psychometric properties: Reliabilities, i.e. Cronbach alphas for each of the four-item facets.

    4. Descriptives for JSS facets for SEP (use 2 decimal places).

    5. Statistical test to compare Sedalia facet means to the JSS norms for Manufacturing Sector (available on line).

    6. Group differences by job2 to see if non-exempt differ from managers.

    7. Optional: item differences by job2 (where facet means differ).

    8. Optional: repeated measures to determine if SEP means differ among facets.


    Note #1: Be sure to use Academic tables and figures, numbering and label appropriately

    Note #2: Comparing the means of the subscales (i.e. salary, promotion, supervision etc.) may be helpful to get an overall picture of how SEP compares to other companies. If the scores for a particular subscale differ greatly from the private sector norms it may warrant further examination and explanation.

    Note #3. The job satisfaction scores can also be compared by Job Title. The job titles have been coded as following 1 = Non-exempt workers, 2 = Managers.  If you were to use GLM: Multivariate Analysis (MANOVA) to analyze the group differences for each facet, you would use the variable job2 as the factor and the facet subscale scores you computed as the dependent variables.


    Note #4:  If you want to compare mean differences among facets, you would need to use GLM Repeated Measures.


    Good luck!

    Case 2: First Federal


    Note: the current year is 1974


    After a long period of deliberation First Federal president Gene Rice issued the following statement: “This has been a difficult year and First Federal has not met its profit objectives. In contrast, 1973 was a very successful year and many of you received substantial bonuses as a result of our Management by Objectives (MBO) program. The granting of yearly bonuses to branch managers is contingent on the attainment of our fiscal goals. As a result, this year we will be unable to grant bonuses in the same manner as the previous year. When First Federal experienced growth we shared our success with our managers through the bonus system; unfortunately this system also asks our managers to forgo additional benefits when we are all faced with financial challenges. The management of First Federal recognizes that there are many forces contributing to profit losses this year including market pressures, and we appreciate the fact that our managers have demonstrated great commitment to First Federal. We are currently evaluating the MBO program, reviewing individual performance ratings, and developing a new strategy for rewarding excellence.”

    Later on that week, the management of First Federal Savings polled its branch managers regarding their attitudes toward the news that they would not be receiving the bonuses they came to expect. The poll asked the employee to rate their level of agreement with a statement reflecting his or her attitude toward not receiving a bonus in the coming year. 


    The "nobonus statement":  (Var Name= nobonus)


    1. "With regard to the decision to withhold bonuses for this year I ..."


    Scale ranged from: 1. "strongly disagree with the decision to withhold" to  5. "strongly agree with the decision to withhold"


    This item was included in a survey of employee attitudes along (1) The Specter Job Satisfaction survey (JSS) with which you are familiar, (2) Six items measuring Intrinsic Motivation, and Six items measuring Growth Need Strength (GNS) (see below for individual items). 

    Intrinsic Motivation Items (Var Names: IM1 to IM6)

    You should compute the sum of The Intrinsic Motivation (IM) items (imtotal) which will result in a possible range of 6 to 36. 

    1= Disagree Very Much; 2= Disagree Somewhat; 3 = Disagree Slightly 4 =Agree Slightly; 5 =  Agree Somewhat, 6 = Agree Very Much

    IM1. I feel a sense of personal satisfaction when I do this job well.
    IM2. My opinion of myself goes down when I do this job badly.
    IM3. I take pride in doing my job as well as I can.
    IM4. I feel unhappy when my work is not up to my usual standard.
    IM5. I like to look back on the day's work with a sense of a job well done.
    IM6.  I try to think of ways of doing my job effectively.

    Warr, P. Cook, J. & Wall, T. (1979). Scales for the measurement of some work attitudes and aspects of psychological well-being. Journal of Occupational psychology, 52, 129-148.


    IM Norms: (two samples:)


    Mean and SD of 18 (5.45)

    Alphas .82 (single factor from Principal components)


    Growth Need Strength Scale (Var Names: GNS1 to GNS) Growth Need Strength items are labeled: GNS1 to GNS6

     You should compute an average (mean) of the six  GNS  for a GNS score with a possible range of 1 to 5.

    1= Not at all; 2= Somewhat; 3= Moderately; 4= A lot; 5 = Extremely

    For the following 6 questions indicate the degree to which you would like to have each characteristic present in your job:

    GNS1. Stimulating and challenging work.
    GNS2. Chances to exercise independent thought and action in my job. 
    GNS3. Opportunities to learn new things from my work. 
    GNS4. Opportunities to be creative and imaginative in my work. 
    GNS5. Opportunities for personal growth and development in my job. 
    GNS6. A sense of worthwhile accomplishment in my work.

    GNS Norms:

    GNS Mean (SD) for managerial job family:  3.3 (.54)

    Hackman, J. & Oldham, G.R. (1980) Job Characteristics Model. Job Diagnostic Survey (In Work Redesign, Addison Wesley Pub.) 

    Analyze these data and create a report to Gene Rice on the potential impact of his decision regarding bonuses.

    (1.) What kinds of branch managers would be likely to give a more or less negative response to the bonus announcement?

    (2.)  How do these branch managers' scores compare to the JSS private sector norms?
    What are the long-term ramifications of the elimination of 1974 bonuses?

                                                                                    CASE 3: MEGALITH

    The Case of the disappearing financial managers

    Because of John Boyd’s concern that the top financial managers at Megalith were leaving due to inadequate compensation, consultants from Hay Associates were called in to assess the internal and external pay equity for these managers. Exhibit 8 in the case shows the salary survey data. Still not convinced that pay was a problem, Frank Nicodemus, VP of HR, called in another consultant group, Personnel and Organizational Psychologists, Incorporated (POP, Inc.) to get another opinion on the matter of turnover among financial managers. This organization has established an excellent reputation as a problem solver for organizational issues ever since they devised their slogan, “You can always trust POP.”


    As psychologists, POP consultants naturally looked to psychological issues. They decided to assess the nature of the management folks at Megalith to determine the relative importance that external rewards and intrinsic rewards played in the motivation to stay. Relying on Deci and Ryan’s SDT Theory, they assessed 70 managers on the General Causality Orientation scale to get a better sense of where these managers stood on the three dimensions measured by this scale.


    Psychometric information on this measure can be obtained from:


    General Causality Scale (Deci & Ryan)

    See also: Deci and Ryan's site for Self-Determination theory
    The POP consultants matched scores for 70 top managers with their current pay scales to determine if pay was related to any of the three dimensions of GCOS. They hoped that results from a data analysis would shed some light on the turnover issue. The data were made available about a week ago to POP junior consultant teams working on the case. Megalith management is expecting a report soon on  the issue.

    Norms available for comparing Megalith sample means on the three dimensions of the GCOS are from a sample of private sector managers:

    Scale*            N        Min    Max    Mean    SD

    Autonomy    59        40        84    68.81    8.85

    Control        59        30        66     48.64   8.20  
    Impersonal   59        16        54     33.66   8.77
    * 12 vignette version (revised)

    You must figure out what research questions / hypotheses can be explored given the existing data available.

    And conduct the analyses to test the hypotheses that you formulate. Some suggested questions to explore are:

    1. Do GCOS dimension means differ by pay category? (MANOVA with univariates and Tukey post hoc)
    2. Do GCOS Megalith Means for Autonomy, Control, and Impersonal, differ from norms of private sector managers? One sample t-tests (don't forget to report means)
    3. Do GCOS Megalith Means for each pay category differ from norms of private sector managers? Hint: with SPSS use Data/Split file to run the one-sample t-test for all four pay categories at once.
    4. Based on the results, what would be some useful interventions and further questions to explore? How would you explore them? What are some possible theories (approaches) that would apply here (motivational, satisfaction, leadership)?

    Management is expecting a written report and Team PowerPoint presentation (for one of the teams) as soon as the results are in.  Good luck.