DataCamp (online data science courses)
SAGE: Which Stats Test? (wizard to determine which statistical test to use when analyzing data)
Stat Trek (statistics tutorials)
UCLA Institute for Digital Research and Education (resources for teaching and learning statistics): annotated output | data analysis examples | faqs | textbook examples | which statistical test?
Hypothesis testing: BC Open Textbooks
Bivariate/multivariate analysis of categorical variables
Chi-square: Stat Trek
Bivariate/multivariate analysis with categorical IV and continuous DV
Bivariate/multivariate analysis of continuous variables
Other topics
Fixed-effects models: SAGE (log in via VCU Libraries)
Matching: Matching Methods (Duke)
SPSS Cheat Sheet: Step-by-step instructions for doing various statistical analysis procedures in SPSS. A Creative Commons-licensed worksheet by Victor Tan Chen.
Kent State SPSS Tutorials | recoding variables | crosstabs | chi-square | correlation | ANOVA
Other SPSS courses: Atomic Learning | LinkedIn Learning (access through VCU) | Lynda.com | YouTube (search for specific SPSS procedures)
SPSS 29 Brief Guide (in Windows, the sample files to use for the tutorial can be found here: Program Files > IBM > SPSS Statistics > Samples > English)
The SPSS Brief Guide (linked above) walks you through the process of opening a dataset, but here is a summary of those instructions with some additional points to consider. If your dataset is in the normal SPSS Statistics file format (*.sav), then you should be able to: (1) double-click on the file to open it; or (2) launch SPSS and, once the intro window opens up, click on the Recent Files tab at the bottom left and choose the Open another file option; or (3) from the Data Editor or Output windows, choose the File > Open > Data option.
Sometimes, you might not be able to find a dataset in the SPSS file format. The data may only be available in a format for other kinds of statistical analysis programs, such as Stata (.dta) or SAS (.sas7bdat). Or, the data may be offered in a simple text format (.txt, .dat, .csv, .tab) or a portable file format (.por), which are more universal file formats. To open these files, choose either option (2) or (3) above. Note that you must extract a compressed (zipped) file before it will be visible using option (2) or (3). On a PC, you can right-click the file and choose Extract All; on a Mac, double-click on the file to extract it. IMPORTANT: When browsing for the data file on a PC, you will need to switch the Files of Type drop-down menu at the bottom from SPSS Statistics (*.sav, *.zsav) to the correct file type: e.g., Stata (*.dta), SAS (*.sas7bdat), Text (*.txt, *.dat, *.csv, *.tab), or Portable (*.por).
Once you find the data file on your computer, select it and click Open. If you don't see the file where you put it on your computer, make sure you extracted the zip file so that it is visible to SPSS. If you are using a PC, also double-check that you selected the right Files of type option (as described above) so that the file is visible.
There are additional instructions in Chapter 2 of the SPSS Statistics Brief Guide (linked above) for opening and converting datasets that are text or Excel files and stored in other file formats.
Go to the General Social Survey page. Information about the survey's methodology can be found on the FAQ page linked at the bottom. Note the discussion of the sample (Which population did the GSS target?) and appropriate weights to use (Do I need to use weights when analyzing GSS data?).
To download the 2018 GSS data, go to the Data link in the top menu. Click on SPSS. Under Individual Year Data Sets (cross-section only), click on 2018. The compressed (zipped) file will be downloaded to your computer. Extract the zip file: on a PC, right-click and choose Extract All, and on a Mac, just double-click the file. Then you can open up the data file by (1) double-clicking on it within the folder you just extracted, or by (2) following the instructions above for Opening a Dataset (e.g, using the File > Open > Data command while within the SPSS program).
To download the 2018 GSS questionnaire, go to the Documentation link in the top menu. In the sidebar on the right-hand side of the screen, click on GSS Questionnaires. Click on 2018 Questionnaires. Under English Questionnaires, you can choose between three different versions of the questionnaire, called "ballots." (The GSS has a split-ballot design, whereby each respondent is randomly assigned a particular version of the questionnaire, with different topics covered in some sections of the survey interview.) Click on the ballot you want to use, and you will download the PDF of the questionnaire.
In the SPSS cheat sheet (linked above), follow the instructions for Applying weights. The variable to use for the GSS 2018 is WTSSALL.
In the SPSS cheat sheet, follow the instructions in the Univariate Statistics section for Frequency tables. You can choose any categorical variable for your analysis (e.g., NATENVIY). Note that certain response categories are listed as missing variables, which means they are excluded from your analysis; see the Setting value labels and missing values instructions in the cheat sheet for more information.
In the SPSS cheat sheet, follow the instructions in the Univariate Statistics section for Bar Charts, using the same variable that you used in the previous step. Follow the instructions there regarding copying and pasting the table you generated into a word processor program like Google Docs or Microsoft Word or a presentation program like Google Slides or Microsoft PowerPoint.
In the SPSS cheat sheet, follow the instructions for Saving output or dataset. Definitely save your output file. Save your dataset if you wish to continue to apply the same weights the next time you conduct your analysis; otherwise, you will need to repeat this step every time you use SPSS.
In the SPSS cheat sheet, follow the instructions in the Univariate Statistics section for Other Descriptive Statistics Tables. You can choose any scale-level variable for your analysis (e.g., AGE). Note that scale-level variables will have a ruler icon next to them when you are viewing variables.
In the SPSS cheat sheet, follow the instructions in the Univariate Statistics section for Histograms, using the same variable that you used in the previous step. As noted in the instructions, it is possible to change the binning options to make a histogram with the degree of precision you wish.
In the SPSS cheat sheet, follow the instructions in the Bivariate Statistics section for Crosstabs. You can choose any two categorical variables for your analysis (e.g., SEX as your IV, SPKCOM as your DV). Remember that the choice of whether to percentage by columns or rows will dramatically change the interpretation of your results. Make sure to generate the chi-square table as described in the instructions; note that the assumptions of the chi-square test must hold for your results to be valid.
In the SPSS cheat sheet, follow the instructions in the Bivariate Statistics section for Clustered Bar Charts, using the same variables that you used in the previous step.
If one of your variables has so many response categories that the chi-square test's assumptions do not hold (or if you just want to make your data easier to present), you can merge response categories in your data by recoding the data. In the SPSS cheat sheet, follow the instructions for Recoding variables. For example, you could recode the RACECEN1 variable into a RACECEN1_recode variable with just 3 racial/ethnicity response categories rather than 16. In the Old and New Variables pane, you would set the old Value of 1 to the new Value of 1 (keeping white respondents as is), the old Value of 2 to the new Value of 2 (keeping African American respondent as is), the Range of 3 through 16 to the new Value of 3 (putting all other races/ethnicities into a single response category), and All other values on the left-hand side to Copy old value(s) on the right-hand side (keeping the missing values as is). After doing the recoding, you will need to create value labels to match what was previously listed in the original RACECEN1 variable (except for your new response category, 3, which should be relabeled as "Other races/ethnicities" or something similar). Likewise, you should set the 0, 98, and 99 values in the recoded variable to be flagged as missing data, following how they were set up in the original RACECEN1 variable (see the Setting value labels and missing values instructions in the cheat sheet).