Hi Jada,

If you need to test the relationship between the dependent variable and the independent variables, you may use the Logistic Regression.

@Mohammed, in Logistic Regression, the dependent variable is binary.

Regards, Witold

Hello again, 

You could use linear regression, however you would need ot convert your categorical predictor into a dummy variable (dichotomous). I wonder how many answer categories does the predictor have? Is Likert type?

Regards, Witold

Hi Jada,

In this situation, I would have used ANOVA. In ANOVA, independent variables are of nominal scale (categorical example : 3 set of pricing , so level 1,2 , 3) ) and dependent variable is continuous (example: sales) or at least interval scaled.

I have done some experiment, where I have set independent variables are of nominal scale and dependent scales are of continuous (example : sales) or scale ( example: purchase intention in the of 1-5 likert scale)

Hope this helps,

Regards,

Gobinda

I think simple linear regression is fine in this case just define your level of categorical variable

Hi Jada,

I think it depends on number of dependent, independent, and control variables.

Case 1: There is only one continuous dependent variable, there are one plus categorical independent variables, and there is no control variable at all, then you can go for ANOVA.

Case2: There is only one continuous dependent variable, there are one plus categorical independent variables, and there are one plus control variables, then you can go for ANCOVA. 

Case3: There are two or more continuous dependent variables, there are one plus categorical independent variables, and there is no control variable at all, then you can go for MANOVA. 

Case 4: There are two or more continuous dependent variables, there are one plus categorical independent variables, and there are one plus control variables, then you can go for MANCOVA. 

Case5: If there is only one continuous dependent variable, there is only one categorical independent variable (i.e. dichotomous, e.g. pass/fail), and no control variable, then go for t- Test.

Regards,

Sujit

Maybe logstic regression for modelling the data.

@ Shah Limon in case of linear regression, you will need that the dependent and Independent variables to be continuous.

ANOVA or F-test is the most appropriate one

if the independent variable is 2 categories, we can use z-test for mean, T-test, ANOVA and Linear regression

if the independent variable is 2 or more categories, we can use ANOVA and Linear regression

I agree with @ Sujit Kumar Ray

Hi,

You may use probit and logistic (logit model) regressions.

Use ANOVA.

Definitely, the fitting statistical test is ANOVA

ANOVA is the appropriate tool for such kind of datasets

If your categorical variable is of two levels, say true/false, then, apply t-test while ANOVA can be applied if your categorical variable is of level more than two, say, continent for an instant, i.e Europe/Asia/Africa. Nevertheless, suggested cases by @Suji Kumar Ray are still valid.

Please check the following website

https://stats.idre.ucla.edu/other/mult-pkg/whatstat/

Two independent sample t-test

Hi Jada,

it simply depends on the nature (distribution) and the number of the variables that you are using.

From your question, you seem to have one dependent variable (which is continuous) and one independent variable (which is categorical)

Scenario 1

If the dependent variable is normally distributed and you have a categorical independent variable that has just 2 levels (dichotomous) then you use INDEPENDENT T TEST.

However, If the dependent variable is not normally distributed and you have a categorical independent variable that has just 2 levels (dichotomous) then you use MANN WHITNEY U TEST.

Scenario 2

If the dependent variable is normally distributed and you have a categorical independent variable is paired then you use a PAIRED T TEST.

However, If the dependent variable is not normally distributed and you have a categorical independent variable is paired then you use WILCOXON SIGN RANK TEST.

Scenario 3

If the dependent variable is normally distributed and you have a categorical independent variable that has 3 or more levels then you use ONE WAY ANOVA.

However, If the dependent variable is not normally distributed and you have a categorical independent variable that has 3 or more levels then you use KRUSKAL WALLIS TEST.

Scenario 4

If the dependent variable is normally distributed and you have a categorical independent variable with 3 or more repeated measures, then you use a RANOVA TEST.

However, If the dependent variable is not normally distributed and you have a categorical independent variable is paired then you use FREADMAN TEST.

I hope this answers your question. Cheers

In such conditions i.e when all independent variable are categorical and dependent variable is continious it is better to use ANOVA

Hello dear ,

Well if your independent variable has 2 categories then you go with the independent sample t test;

however if your IV has 3+ you go with one way ANOVA.

Dr. M. A. Fakhro did not mention anything abut the dependent variable, therefore, I will not assume here. What if there are 2 continuous dependent variables and one dichotomous independent variable? Do we still go on with Independent Sample Test or with MANOVA?

Umamaheswararao Jada if you have more than one dependent variable you can go with MANOVA

You use ANOVS

You use ANOVA

Independent T test or ANOVA can serve. MANOVA can also serve, if the dependent variable is more than one.