In general, for inference statistics, understand how to establish null and alternative hypotheses from any problem. For instance, what is the null hypothesis of our umbrella topic on Gender Wage Gap? Use that as an example, and create scenarios where you have sample size n=104365, and assume values, such as your average estimate is beta_hat=.45 and your standard errors estimated are equal to 0.10. Use the right statistical table to answer this question.

There is also the part to understand what then result will mean. Are your values of the dependent variable in logarithm terms or level terms? Are you control variables in levels or logarithm terms? This will affect the interpretation of the estimated coefficient.

Further, also understand what rejecting or not rejecting the null hypothesis does to your conclusion in your model. For instance, if in the Gender Wage Gap model our Null hypothesis is beta_sex=0 (Beta coefficient related to the variable sex) meaning there is not a difference between male and females in average wages, then rejecting the null hypothesis means that we find evidence against that hypothesis. So, there is likely some differences and the t-test would allow us to tell at what degree of significance level that difference may exists.