In an intervention study to determine whether a vegetarian diet can help college women to lose weight, 500 residents of a women's dormitory were divided into two groups: one that ate in a vegetarian dining hall and one that ate a normal diet. women were allowed to choose which group they wished to be in. after three months, the vegetarian group had lost an average of 10 pounds while the control group had not lost weight. what is the most likely source of error in this study?
a. reporting bias
b. the placebo effect
c. random variation
d. selection bias