So for the proportion, for example person 1 had. HI Karen, I have two variables – one is nominal (with 3-5 categories) and one is a proportion. Are they measuring independent or dependent variables? What OTHER variables are you using in your analyses? Finally, I very much doubt that I’d do much analysis of individual items from your survey (if they are subsumed in your scales). (With four scales, you’d have six correlation coefficients to examine the correlations between the six pairs!) Then, to your logit regression question: Not sure that I know how you’re treating these four scales. So, I’d construct a simple Pearson product moment correlation matrix to examine the correlations between each of the pairs of scales. Now, I’m also always concerned about these scales’ “independence” from each other. What do the scales MEASURE? That is, what variable/construct/concept does each scale quantify? I’d produce descriptive statistics to describe each of the scales/results from the summing. (You might use something like Cronbach’s alpha to provide you some evidence. Mmm, first, I’d wanna know how INTERNALLY CONSISTENT each of the summed scale scores was. So even in a very simple, bivariate model, if you want to explicitly define a dependent variable, and make predictions, a logistic regression is appropriate. So the question is, do you want to describe the strength of a relationship or do you want to model the determinants of and predict the likelihood of an outcome? It’s not a modeling technique, so there is no dependent variable. Why not just use the simplest of all?Ī Chi-square test is really a descriptive test, akin to a correlation. My personal philosophy is that if two tools are both reasonable, and one is so obtuse your audience won’t understand it, go with the easier one.
So, if given the choice, I will use logistic regression. That said, I personally have never found log-linear models intuitive to use or interpret. A log-linear analysis is an extension of Chi-square. If all the variables, predictors and outcomes, are categorical, a log-linear analysis is the best tool. Now you could debate that logistic regression isn’t the best tool. Simpson’s paradox, in which a relationship reverses itself without the proper controls, really does happen. Including controls truly is important in many relationships. And whether any of this is actually true, I’m sure people worry about it. Mainly because it communicates (on some level) that you understand sophisticated statistics, and have checked out the control variables, so there’s no need for reviewers to object. I’m sure there is a bias among researchers to go complicated because even when journals say they want simple, the fancy stuff is so shiny and pretty and gets accepted more. Of course I can’t say why anyone uses any particular methodology in any particular study without seeing it, but I can guess at some reasons. You’re right that there are many situations in which a sophisticated (and complicated) approach and a simple approach both work equally well, and all else being equal, simple is better. Per your question, there are a number of different reasons I’ve seen. I look forward to seeing you on the webinars. I enjoy reading your site and plan to begin participating in your webinars. My professors don’t seem to be able to give me a simple justifiedĪnswer, so I thought I’d ask you.
It also just seems so much more simple to do chi-square when you are doing primarily categorical analysis. I have worked with some professionals that say simple is better, and that using Chi- Square is just fine, but I have worked with other professors that insist on building models. I read a lot of studies in my graduate school studies, and it seems like half of the studies use Chi-Square to test for association between variables, and the other half, who just seem to be trying to be fancy, conduct some complicated regression-adjusted for-controlled by- model. Why is using regression, or logistic regression “better” than doing bivariate analysis such as Chi-square? I am an MPH student in biostatistics and I am curious about using regression for tests of associations in applied statistical analysis. I recently received this email, which I thought was a great question, and one of wider interest…
0 kommentar(er)
