Learning two level multilevel regression: A jamovi-based approach (video 2)
HTML-код
- Опубликовано: 22 авг 2024
- This video is the second of 2 in the series (the first being: • Learning two level mul... ) aimed at providing an introduction to two-level multilevel regression modeling using data from the High School and Beyond Survey from the year 1982. I demonstrate concepts and strategy using the jamovi program, which you can download from here (www.jamovi.org.... The examples in the video involve predicting student math achievement as a function of student level ('ses') and school-level ('mean_ses', 'schtype', 'size') variables.
I walk you through (a) a model that includes 3 level 2 predictors (instead of the one from the previous video), (b) a model where the level 1 is allowed to randomly vary, (c) models including tests of cross level interactions. I also demonstrate how you can probe cross level interactions. You can obtain a copy of the data here (drive.google.c.... You can download a copy of the Powerpoint from this second video in the series here (drive.google.c....
For more examples of multilevel modeling using various programs, go here: sites.google.c...
For those interested in learning multilevel modeling with SPSS, there is a nice workbook that can be found here: www.routledge....
Thank you for this! Very useful. :)
Hi MIke, Thank you for the video. To run a cross-level interaction (level 1 predictor and level 2 predictor), could you alternatively click on the number before dragging the term into the Model Terms field?
Mike, how can one use jamovi to (statistically) compare the 5 models discussed in your video series? Is the simple visual inspection of AIC/BIC/loglike sufficient? Should one use the chi square after manually subtracting the corresponding values for each model?
Hi Mihai, your question prompted me to create an Excel calculator (drive.google.com/open?id=1kRZcpJ7orjlIBU8790Eb79-8YsGoGnp1) for performing a chi-square difference test, along with this demo (ruclips.net/video/bzqMFnl3IV4/видео.html). Hope you find these useful!
@@mikecrowson2462 Would be kind to make the two links live again? Thank you!