Employee dataset

PowerPoint for Regression Analyses

Powerpoint for Factor Analysis

Assignments for PSY6010 include weekly readings, with a statistical problem, which must be performed and results interpreted. SPSS Syntax and output must be provided. Please copy the spss table and paste it into the word document as a picture.

Assignments

• Due February 4

  Using the Employee Survey, do the following exercises.
  
  1.  Our research question is: What is a good job?  
  	Explain why this is important (1 paragraph).  Provide an example 
  of a client.
  	Select a measure of job satisfaction and explain why you think 
  it.s a good one. (1 paragraph)
  	Provide a set of factors that you think would determine job 
  satisfaction.  First discuss them as concepts, and then discuss the 
  measurements of them.   (1-2 paragraphs)
  
  2.  Using your measure of job satisfaction:
  	Run a frequencies test, with mean and standard deviation.
  	Run a frequencies test on all the IVs.  
  
  	Explain what your results mean for your potential analysis: will 
  these variables work?
  
  3. Do a compare means for your dependent variable and a dichotomous IV. 
  Check the anova option.  Then run a One-Way Anova (from Compare Means)  
  Explain your results.  
  
  4.  Run a correlation of your dependent variable and 3 other independent 
  variables. Explain your results.
  
  5.	Try a simple regression:  use only interval or continuous 
  variables as IVs for this exercise. Explain your results.  What do you 
  conclude from your simple model?  How might you go about improving it?  
  
  Include your syntax and output with your explanation.  
  

• Due February 11
  Excel spreadsheet for class Second spreadsheet for class

  
  Using the Employee Survey, do the following exercises.
  
  1.	Run a crosstab of overall job satisfaction and .challenged and 
  interested in my work..  Are there any logical outliers?  Describe and 
  explain how you might go about handling them.  
  2.	For the missings in overall job satisfaction, replace them with 
  the mean, and create a dummy variable for missing.  
  3.	Change the string variable for race into a dummy variable for 
  whites versus everyone else.  
  4.	Change age (v54) into a numeric variable. Recode the values to the 
  midpoint value.
  5.	Run a compare means for sufficient opportunity for career 
  advancement by How long you.ve worked in this company (Tenure).  Conduct a 
  linearity check as well.  Would you use tenure as an explanatory variable 
  for career advancement perception? Why or why not?  
  6.	Now try this with hours worked each week instead of tenure.  Would 
  you use hours worked as an explanatory variable?  Now transform hours 
  worked to 51 hours or more versus less.  Explain why you might do this.  
  Run the compare means again. What are the differences and what does that 
  tell you about hours worked and career advancement attitudes?
  7.	Run a correlation of v20-v25.  Is there evidence of 
  multicollinearity?  How would you address this if using these variables to 
  explain overall job satisfaction? What would you do and then, make the 
  variable changes, if any.
  
  
  Extra credit:  Calculate a variable for how long it took to complete the 
  survey (start date versus end date).  What are the outliers?  
  
  Include your syntax and output with your explanation.  
  
   

• Due February 25
  PowerPoint for Regression Class

  Using the Employee Survey, do the following exercises.
  
  1. Choose a dependent variable that is amenable to be used as a 
  dependent variable in a regression.
  2. Select two other continuous, interval or ratio variables (scale) 
  as independents, and one dichotomous variable.
  3. If the correlation is above .65 for any independent variable, 
  choose another.  
  4. Explain your research question as done in the PowerPoint, develop 
  the model and hypotheses, and write up your methods section.
  5. Run the regression using the standard method. 
  6. Write up your results and interpretation.  
  7. Calculate two different levels of your dependent variable by
  slotting in different values of the IVs.
  
  Include your syntax and output with your explanation.  
  

• Due March 3

  1. Create an interaction term by computing a new variable COMPUTE NEWVAR = oldvar1 * oldvar2.
  2. Run an OLS regression where Y = your dependent variable (continuous, scale or interval) and x1 = one variable, 
  x2 = the second.   y = a + b1x1 + b2x2.
  3. Now run the regression adding in the interaction term as well interaction = x1*x2.   
      y = a + b1x1 + b2x2 + b3interaction.  
  4. interpret the differences in the two models.  
  

• Due March 10

  Using the Employee Survey, do the following exercises.
  
  1.	Using one of the non-categorical variables as the dependent 
  variable.  Create a conceptual and hypothesized model, explaining your 
  variables and method.
  2.	Run a regression with two dichotomous variables and one ratio or 
  interval variable.
  3.	Calculate an interaction term with the two dichotomous variable 
  and rerun the regression, providing logic for the interaction. Did it 
  significantly contribute to the model?  Explain what the result means.
  4.	Now calculate an interaction term for the dichotomous and the 
  non-categorical variable, providing logic for this interaction.  Run the 
  regression. Explain the results. 
  
  Include your syntax and output with your explanation.  
  

• ALSO FOR March 10

  Your homework is to do the assigned reading re: logistic 
  regression (chapter 6A and 6B)
  
  Begin to use your  dataset for your 
  project by downloading a dataset or obtaining the data from me,  and 
  submitting the codebook to me for the 
  assignment.  Here is a link to instructions
   for working with a data set from the SDA archive.  
  
  If you don't have another data set at this point, then you MUST use the GSS.  
  However, if you do, then it's time to show me what you have and by that 
  I mean a codebook with lists the variables you will use in your models 
  (proposed), the question text, the number of cases, and frequencies for all 
  values.  
  
  DUE March 17: 
  Using our Employee Survey, do the following exercises.
  A.
  1.	Identify a research question you want to explore where the DV will 
  be a dichotomous variable.  Identify 2 dichotomous independent and 1 
  interval independent variable, and state your hypotheses.  
  2.	Conduct diagnostics.chi-square tests for the dichotomous IVs and 
  compare means for your interval IV.  Are your hypotheses supported?
  3.	Run a logistic regression with your variable and 2 dichotomous 
  independent and 1 interval independent variable
  4.	Explain your results:  the B coefficients, the Exp(B) (odds 
  ratios), their statistical significance, and the model fit.
  5.	Interpret these results.  
  
  B.  Using your custom GSS data set, explain your conceptual model and how 
  you.ve operationalized it.  You will need to consider identifying 
  multidimensional scale items for collapsing using either factor analysis, 
  or some other variable transformation (don't worry about FA yet, we 
  haven't covered it.)  
  
  Include your syntax and output with your explanation.  
  
  
  DUE March 24: 
  Using our Employee Survey, or your own dataset, do the following 
  exercises.
  
  1.	Identify a dependent variable suitable for ordinal logistic 
  regression, and one suitable for multinomial logistic regression.  Explain 
  briefly. 
  2.	Conduct diagnostics, that is, crosstabs with the chi-square tests 
  for the dichotomous IVs and compare means for your interval IV with the 
  anova.  Do we see evidence of group definition?
  3.	Run the ordinal logistic regression and multinomial logistic 
  regression for your respective DVs.  Use the same independent variables if 
  you wish.  Explain your results:  
  4.	Interpret these results.  
  
  Include your syntax and output with your explanation.  
  
  
  
   March 31 SPRING BREAK!!
  
   DUE April 7 Start using your database.  
  
  
  1.  Run the frequencies for all the variables you'll be using in your model.  
  
  2.  Run a correlation of your dv(s) with your iv's.  
  
  3.  Try out a very simplistic version of your model. 
  
  4.  Describe your results.  
  
  
  
  This powerpoint will be used in class on April 7 and April 14:  Powerpoint for Factor Analysis
  
  
  DUE April 14
  
  
  Homework #8 on Slide 14.  Run a frequencies test to make sure they are appropriate.  Are they? 
  Explain.
  Run a correlations table.  Is this appropriate for PCA/FA? Explain.
  On this same selection of variables, conduct tests for KMO and 
  Bartlett.  Are we still on track for PCA/FA?  Explain.  
  Now conduct a factor analysis using these variables, setting the 
  defaults as in the class example. Are you happy with this result?  Then 
  try setting  the number of components differently, adding one or more, or 
  subtracting, from the first result.  Are you happy with this result?  Explain.  
  What can you say about components of Quality of Work?
  
  
  
  DUE April 21
  
  
  See  Analyses for Experimental Designs.  
  
  Using the same factor analysis you ran last week with the employee 
  data 
  See slide #14, run this factor analysis and save the factor score 
  variables.
  Now run a regression:
  Overall satisfaction = a + (factor scores) + male + hours worked 
  (hourswk)+ whether there was a layoff (v32)
  Explain why this model makes theoretical sense.  Now explain the 
  results.  
  If you were an HR manager, what areas would you either try to improve, 
  or 
  make sure they stay as good?
  

Also for April 21 - The MIDTERM!

Materials for Midterm
• Review document for midterm
• Dataset to be used on midterm
• Codebook for the midterm dataset.