Discussion

I used several techniques to minimize the bias and error in my study. First, I recorded data for every customer that passed through my line, negating the need for a sampling technique. This also gave me a very large sample of 120 for women and a generous 40 points of data for men. I also eliminated the confounding variable of coupon use by recording totals before coupons were applied, which I noticed was much more prevalent among women than men. One factor that may have affected my data is the time I collected my data, which was only on weekend afternoons. As my background research indicated, time of day and time of week have a significant effect on who is shopping, and how much they are spending. Larger studies that I referenced in my background research indicated that there is a small, but measurable difference between the average grand totals between men and women. If I used a larger sample over a variety of different days of the week, I believe I could have concluded there is a difference between the grand totals of men and women.