GSS Database:;jsessionid=EB648CC32E7868756B6A741103078A83?dataset=gss14nw

Find variables that are applicable to the workplace:  For example, the folder “1989 ISSP MODULE: WORK ORIENTATION” contains many variables that are suitable for this project.  Other folders have interesting work-related variables as well.  Do some searching and explore the database. 


  • Insert your Theoretical Model – Figure 1:

Give your model a descriptive name and add it to the text.  Let your model take up a whole page if needed. 

  • Check to see if you get Correlations:

Using the “Correl. matrix” tab on the right-hand side of the database, enter all of your variables by selecting your variable in the folder it is contained in and in the “Variable Selection” area, Copy the variable to the “Correl. matrix” area by clicking “Vars to Correlate”…you will see it move over to the “Variables to Correlate” list…the grid allows for 8 variables at a time.  Once all your variables are in the grid, click the “Run Correlations” button at the bottom of the screen. 

  • Craft logical Hypotheses:

Write them out based on the positive or negative relationship you think the two variables will have.  Develop a hypothesis for each predictor with the outcome.  This is challenging, but fun.  You may find you have to re-write them, but keep it simple to start with.  Students may also find that a predictor they initially thought made logical sense, does not make sense.  That is OK.  You can search the GSS folder you utilized for your other variables for a new one – again, these models can and should be refined during this process.

It may also be the case that you end up with (or seek out) Control variables – things that rule out alternative explanations for your hypotheses (for example, in my Job Embeddedness 🡪 Work stress article, I controlled for Job Satisfaction so I could say my results hold true above and beyond an employee’s attitude toward their job).  Having control variables is not mandatory, but sometimes researchers end up with them or it is convenient to measure them and use them later (like demographics).

  • Search, download, read and summarize relevant research articles

No less than 10 empirical articles to be included in your project. Keywords are critical! Conduct key word searches on Google Scholar  – there is much published on your topics & variables. 

Start by reading the Abstract & Discussion section.  This will give you a good summary of the article and let you know how it fits with your hypothesis development.  That is, you must craft strong arguments as to why you believe your hypotheses are correct.  What supporting evidence can you show the reader to convince them your logic and reasoning is sound and supported by the literature. 

  • Start developing and organizing your Methods section:

In Step 1, it was suggested that you extract as much information on your variables of interest – this is helpful for several reasons.  Attempt to assess how your predictors & outcome are measured in the literature that you researched and downloaded.  Write down the name of all of your variables as well as the phrasing of the measure – how is it worded?  What scale was used?  What name did you give the variable (e.g., RECGN = Recognition in the workplace).

To set up your Methods section, review the Job Embeddedness paper provided and also use one of the empirical articles you have downloaded.  There are different ways to organize a Methods section, but you will notice some similarities – Authors need to tell the reader as much as they can about the Procedures, Sample, Measures, Data analysis (we will get to that)…but students will notice that the more information provided, the better. 

Explore the GSS website for information regarding descriptions of the sample, procedure & data collection. Alternatively, search (Google Scholar) for empirical articles that have used the GSS and use that as a template to organize your Methods section.


  • Check to see if you get Correlations:

If an error occurs on the message, the data you are analyzing likely does not match (for some reason) and you may have to test one predictor at a time (with your outcome variable) to see which one is ‘problem’…explore the folder of origin for useful variables – and ask for help if needed.

  • Hypotheses

Each predictor will have a hypothesis with your outcome.  You have to propose what direction (positive or negative) the relationship will be, with the assumption that will be statistically significant.

EXAMPLE: As (perceptions of) coworker relationships improve and are appreciated, job satisfaction increases (goes up).

ALTERNATIVE: There is a positive relationship between Coworker Relationships and Job Satisfaction.

Check the language and Likert scale of each variable (predictor🡪outcome) to make sure you word your hypotheses correctly – this may be difficult, but again the learning occurs during the struggle. 

  • Start your theory building:

This is the ‘meat’ of the paper and demonstrates to the reader that you did your homework and understand (some of) the literature surrounding your variables.  This is the text before each hypothesis which convinces the reader that your hypothesis makes sense and has some logical arguments and research support/evidence. 

  • Start your Reference section:

Keep track of your research articles and develop a proper reference section using APA style.  This is easily done using Google Scholar and also by using an example article (like my Job Embeddedness study) and copying what you see in certain detail.

  • Run Regression model:

Type of service: Academic paper writing
Type of assignment: Project
Subject: Management
Pages/words: 14/3850
Number of Sources: 10
Academic level: Doctoral
Paper Format: APA
Line spacing: Double
Language style: US English

Why us

Free features

get started