Hi everyone,
I am researching the effect of personality on the perception on advertisement. Every participant ranked the ads on a 5-point likert scale to indicate if they "liked" them. I want to find out if their personality (scores go from 24-120 numerical) correlate with the "liking" (results of likert-scale) of the ads. Each personality type also has 6 (highly correlated) subsets. I am not a proffessional in statistics, but i figuered to use ordinal regression for the main perspnality types, but that probably would not work for the subsets because of multicolinearity.
Is ordinal regression applicable or is there a better way to do it (e.g. Pearson’s correlation)? Also for the subsets I am a bit lost -please help.
Sample size is 230. After preparing the data is suspect that the correlation is not linear.
Linear regression analysis or if you code: do not like = 0 and like = 1
you can use logistic regression analysis. Forget about linear correlations and check the assumptions for a significant ANOVA in your model in the first case.
Hello Fabian Wittmann
if the dependent variable is a likerts score or a sum of scores, then your model is a proportional odd (ordinal) logistic regression.
Multi-collinear variables can be reduced by using an optimum scaling procedure.
https://vuorre.netlify.app/publication/2019/03/01/ordinal-regression-models-in-psychology-a-tutorial/burkner-ordinal-regression-models-2019.pdf
Hi, Fabian,
Multiple regression is the legacy method which would typically be used in your application to maximize explained variation in the ordered personality dependent measure (logistic regression would be the legacy choice if the personality measure is binary and the objective is to maximize likelihood).
No regression model is formulated to identify the most accurate model, and neither variance or likelihood are synonymous with accuracy. It is therefore not surprising that regression-based models cannot make accurate predictions of data values which deviate from the mean (ordered data) or the prevalent (categorical data) response.
- https://odajournal.com/2013/09/20/maximizing-the-accuracy-of-multiple-regression-models-using-unioda-regression-away-from-the-mean/
- https://odajournal.com/2013/10/21/assessing-technician-nurse-and-doctor-ratings-as-predictors-of-overall-satisfaction-of-emergency-room-patients-a-maximum-accuracy-multiple-regression-analysis/
- Yarnold, P.R., Hart, L.A., & Soltysik, R.C. (1994). Optimizing the classification performance of logistic regression and Fisher’s discriminant analyses. Educational and Psychological Measurement, 54, 73-85.
If accurate modeling of your data is important in your application, then there is a modern non-parametric alternative which identifies the most accurate (weighted) model(s) which exist for your data and hypotheses.
- https://odajournal.com/2014/04/06/a-statistical-guide-for-the-ethically-perplexed-chapter-4-panter-sterba-handbook-of-ethics-in-quantitative-methodology-routledge-2011-clarifying-disorientation-regarding/
- https://odajournal.com/2020/06/28/what-is-novometric-data-analysis/
- https://odajournal.com/2016/09/19/novometric-analysis-with-ordered-class-variables-the-optimal-alternative-to-linear-regression-analysis/
***
PS: If you use legacy multiple regression, you may wish to assess whether a non-linear model is appropriate:
- https://odajournal.com/2019/01/19/when-to-evaluate-a-nonlinear-model/
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Regression