I have a dataset in which risk appetite is measured through likert scales and a series of questions in which a choice must be made between sure win and bet (example: do you prefer to get \$ 1000 safe or have a 50% chance of getting \$ 2000?). I have some data of those who responded, including gender. I was asked to calculate the correlation between gender and risk appetite.
Now, to summarize the risk appetite, I thought of calculating the average of the likert scales (a low value equals a reduced risk appetite and vice versa), in order to reduce all the questions to a single variable and to be able to calculate the correlation with the gender. Then I could do the same thing for the answers on sure winnings or bets, reduce them to a single variable by calculating the average, and then calculate the correlation with gender. My problem is that I cannot understand the usefulness of this question, ie what is the point of calculating the correlation between a variable and a dummy variable? Do you think it makes sense to make such a calculation? Could it be that whoever left me with this assignment misused the term "correlation"? Also, I just can't understand how such a value can be interpreted. Assuming that the correlation takes a value of -0.2, what could I say? That men are on average more risk-prone?