I want to know whether the dependent variable should have the same sample size as the independent variable or not.
For example, if we want to study the impact of accountants' education (independent variable) on corporate tax evasion (dependent variable),
let's assume we collected data for the independent variable (accountants' education) by distributing surveys to 100 accountants. As for the dependent variable (corporate tax evasion), we collect it by performing some calculations on the financial statements of 56 companies.
Is this statistically valid?
Thank you.
Hi Tihana Škrinjarić
I was considering inputting the independent variable (x) as a Likert scale and the dependent variable (y) as values in SPSS, which will automatically handle the missing values . and From there, it will be easy to conduct statistical tests like correlation, regression, etc.
However, I then realized that this might be statistically invalid.
What do you think?
thank you.
It sounds like you are trying to correlate two separate samples. Nor do those sample sizes appear sufficient for making generalised conclusions on the population of accountants nor companies. You might be better taking a deep dive case study approach on organizations known to have evaded tax.
yes Christopher C Kelly that's also what crossed my mind, especially the dissimilarity in the nature and size of the data between the two variables.
I was not sure about it, but I'm glad that I asked this question.
Thank you very much.
The sample size for the dependent variable does not have to be the same as the independent variable. However, it is recommended to have equal sample sizes for each group of the independent variable. This is because having unequal sample sizes can reduce the statistical power of your test.
