A cross-sectional study of the effects of participating in a program (e.g., job training) may wrongly conclude that the program is effective (e.g., participants achieve higher rates of employment than similar people without training) because of selection bias. Such bias is the consequence of which one or more of the following factors? (Circle all that apply.) c. High nonresponse rates among sample participants b. The failure to conduct a focus group in order to understand how people think about job training programs e. Program administrators permit only those people most likely to succeed to participate in the f. program (a process referred to as creaming ) d. The failure to employ a backward research design a. People who volunteer to participate in a job training program may have unmeasured characteristics (e.g., ambition) that could produce any observed differences (e.g., employment) in outcomes g. Multicollinearity