Research project

Assignment

Each student will complete a research project during the semester. The research project involves: choosing a topic problem and data set from several different data sets available from the instructor, choosing an appropriate statistical test for hypothesis testing for the problem chosen, running the appropriate test(s) using SPSS, interpreting the SPSS output, and writing a paper (roughly five pages) describing the problem and the conclusions from the statistical test results. This project will test students' ability to apply and conduct an inferential statistic to a specific problem of interest. The research report will be written in APA style and laser-printed (or printed with an inkjet printer of similar quality). Blurred printing, smudged printing, or less-than-laser quality printing is unacceptable and will result in a grade of 0. The projects are due on the last day of class; late projects will not be accepted and will be assigned a grade of zero. Additionally, failing to run a spell-check on the assignment will result in losing 10% for each spelling error that would have been caught by a standard spell-checking program.

1) Download the datafile harass.sav

2) The datafile contains the following variables:
  • Hartrain: Some employees were required (via randomly assignment) to attend the sexual harassment awareness training program (1=not trained, 2=trained) in 1998
  • Gender: female = 1, male = 2
  • Age: age of employee at 1999 testing
  • Joblevel: 1 = staff, 2 = lower management, 3 = upper management
  • Yearscom: Number of years employed at the company at 1999 testing
  • Yearsed: Number of years of post high school education
  • Jobsat: Job satisfaction score from questionnaire (higher scores mean greater satisfaction)
  • Lifesat: : Quality of life satisfaction score from questionnaire (higher scores mean greater satisfaction)
  • Har1997: Number of sexual harassment experiences (including those they had experienced or witnessed) during the past year (1997)
  • Har1999: Number of sexual harassment experiences (including those they had experienced or witnessed) during the past year (1999)

3) Answer each of the following research questions. To do this you will need to identify what analysis is appropriate, carry out that analysis, and know how to interpret the output of that analysis. For all tests, assume an alpha level of 0.05.

  • Describe the demographics (age, men/women, & job level) of the sample.
  • Describe the distributions for harassment in both 1997 & 1999.
  • Which variables are significantly correlated with one another? Describe these relationships.
  • Compare the following three models for predicting job satisfaction. For each model, describe which explanatory variables are significant. Overall, which model do believe is the best (explain your rationale).
    • Model 1: Predictors - life satisfaction
    • Model 2: Predictors - life satisfaction, years at company
    • Model 3: Predictors - life satisfaction, years at company, sexual harasment in 1999
  • Is there a significant change in sexual harassment from 1997 to 1999? Explain the relevant descriptive statistics as well as the inferential statistics.
  • Is there a significant change in sexual harassment in 1999 depending on training program? Explain the relevant descriptive statistics as well as the inferential statistics.
  • Is there an effect of job level on sexual harassment in 1999? Using a Tukey test, which groups are statistically different?
  • Is there a significant interaction of gender and training on sexual harassment in 1999? If so, describe the interaction.

4) Write up the results.

    a. You must use full sentences & paragraphs
    (I recommend at least separate paragraphs for each research question).
    b. Feel free to use any graphs, figures, and/or tables to supplement your text (include these on separate sheets of paper, but refer to them (and label them) in the text.
    c. APA style is not required (but is suggested).
    d. Make sure to provide the relevant descriptive statistics as well as the inferential results.
    e. For simplicity, assume an alpha level of 0.05 for all tests (make sure that you state this somewhere).
    f. Remember to report degrees of freedom when appropriate (along with computed test statistics, DON'T just report p-values)
    g. Papers should be typed, double spaced.

If you have any questions, please feel free to contact me at jccutti@ilstu.edu.