I am doing a research on the awareness of digital banking i Odisha. In order to test association among demographic profile (age, gender ,experience ,income, location) and awareness of digital banking what test should I use(ANOVA OR CHI SQUARE TESTS) ? My respondents will be users of digital banking.
It depends on what type each variable is (numeric, ordered categorical, nominal categorical).
And if you are planing on doing just associations for each pair of variables (Awareness and Age, Awareness and Gender) or if you want to construct a single model with Awareness and several other variables.
Hola, primero, si trabajará con variables cuantitativas, una prueba de normalidad, a veces en vez de ANOVA lo que se necesita es la de Kruskal-Wallis.
Luego, debes considerar:
1. Si usarás solo variables cualitativas, entonces debe ser chi-2, por ejemplo género con conocimiento.
2. Si trabajarás con una variable cualitativa POLITÓMICA, como género, estado civil o algo así, para cruzarla con una CUANTITATIVA, entonces debe ser anova o Kruskal-Wallis.
Si tu, quieres cruzar una cualitativa DICOTÓMICA con una CUANTITATIVA, entonces tu prueba es T de student para muestras independientes, o U de Mann-W, si no es paramétrica, saludos
ANOVA is used for comparing means across multiple groups, while the chi-square test is used for assessing associations or independence between categorical variables. The choice depends on whether you're analyzing numerical data or categorical data.
If age, gender and income are continuous variables and the awareness, your response variable, is a 'yes or no' variable, you can use Point Biserial correlation (given that the assumptions are satisfied, and you plan separate tests). If you have all those variables as categorical variables, then Chi-Square, as suggested by others, can be used. If the variable "location" is grouped as rural, urban etc then, a chi-square between that and the awareness can be done.
- Badges
- Science topic
- Similar topics
- Filing