Dear Hayatun Syamilah

If you want to check their correlation, you are suggested to use the Spearman correlation test.

If you want to examine their association, you should define X and Y first.

• If Y is the body weight status, you could use the logistic regression model.
• if Y is the food intake frequency, the next concern is about whether you want to use the overall score or individual item as the outcome variable. You should also check distribution to see if the outcome is normally distributed or not.
• --> If it is normally (nearly normally) distributed, you could use the linear regression model.
• --> if it is not, or you use the individual food items, you may think about classifying them into categorical variables, and a logistic regression model could be an option to use.

Lastly, different other statistical models might be applied, depending on your study design.

Hope this could help.

Best,

Duong,

Dear colleague

please check at RG: Nominal-by-Interval Association Eta, the Correlation Ratio.

regards,

Redimio

You can use the Spearman's if you want to establish the correlation