As far as the effect size in research is very important SPSS can not calculate it directly. I want to use the effect size calculators that find on https://lbecker.uccs.edu/. Can I believe that the calculated effect size is correct? I need comments and advice.
Hello Tola,
What ES metric do you seek, and what type of analysis did you run? There are many conversion options for taking summary statistics and estimating an ES, or even the test statistic itself (e.g., t-, or F-ratio) and the associated df, and converting to an equivalent ES. You'll need to be more specific.
Good luck with your work.
Thank you very much. I want to calculate Effect Size for the Mann Whitney U test.
Hello again Tola,
The obvious one would be to estimate r as the ES. Since most software packages that report Mann-Whitney evaluate significance via the "large-sample" (z) approximation rather than use exact computation, convert the given z as follows:
r = z / sqrt(N)
Where N is the total sample size (n1 + n2). Other rank-based tests might report a chi-square statistic instead, in which case the formula would be: r = sqrt(chi-square/N).
Good luck with your work.
Effect size can be expressed in terms of strength of association (using the r family) or in terms of magnitude of difference (using the d family). SPSS can provide partial eta squared (η2p) in several statistical procedures that employ ANOVA capabilities. For Cohen's d, you could use the commands reported in the IMB SPSS Site, cited below. Besides, you could use interpretation charts that are rendered in most introductory statistical textbooks (e.g., Morgan et al., 2020). You may refer to Lakens (2013) for relevant insights.
IBM Support. (n.d.). Can SPSS produce an estimated Cohen's d value for data with two independent samples? IBM - United States. https://www.ibm.com/support/pages/can-spss-produce-estimated-cohens-d-value-data-two-independent-samples
Lakens, D. (2013). Calculating and reporting effect sizes to facilitate cumulative science: A practical primer for t-tests and ANOVAs. Frontiers in Psychology, 4. https://doi.org/10.3389/fpsyg.2013.00863
Morgan, G. A., Barrett, K. C., Leech, N. L., & Gloeckner, G. W. (2020). IBM SPSS for introductory statistics: Use and interpretation. Routledge. https://www.routledge.com/IBM-SPSS-for-Introductory-Statistics-Use-and-Interpretation-Sixth-Edition/Morgan-Barrett-Leech-Gloeckner/p/book/9781138578210
Good luck,
SPSS v.27 has built in effect size calcuator for parametric tests. Also the webiste Psychometrica is good https://www.psychometrica.de/effect_size.html
- Badges
- Science topic
- Similar topics
- Comment