A: We have retained subjects by being nice and accepting and by developing a relationship with them. An initial home visit to explain the study has been really important. People feel weird about being in psychological studies. Sending letters and talking to them over the phone does not do enough to reassure them. This is where the home visit comes in. We choose people who are not threatening to our sample, i.e. have a blue collar identification. We then train them to make the families feel comfortable and at home with us. We teach them to join with the family a bit by telling a story about themselves or by complimenting the family. We want to make the family feel like these aren't perfect people coming in to criticize.

A: We pay our subjects well ($150 per interview). I think it's inconsiderate to expect families to come in and spend a lot of time (and it's quite tedious) for token pay. It's not realistic. People don't dislike being in a study, but they are busy and some have chaotic lives. So if you want them to come in, you have to pay them well.

A: They receive a couple of weeks of part-time training, during which they engage in role plays. They are trained in a carefully worked out protocol for the initial phone call to set up the home visit. Then we meet with them every week to see how the home visits are going.

A: For our study, the schools agreed to send a letter to families about the study. The letter was in the form of passive consent: if the family did not wish to be contacted, then they had to contact the school to have their name removed from the list. This works much better than saying "if you want to be in the study, call these people." People don't do that. They don't mind being in the study if you contact them but they won't contact you.

A: First, we contact subjects every year so we don't lose them and? the trail doesn't go cold. We get contact information and update it periodically. Also, the subject contact has been working with the families for 14 years; she knows the habits and personality of each participant.

A: Looking back, those who were the most difficult to recruit (those without phones) were the most anti-social boys in the sample. Recruiting the difficult to recruit is very important in order to achieve a distribution of behavior that reflects your population. Those who are most at risk for dropping out generally are the ones you most want to retain. They are the most anti-social and have the most problems. It is critical to pull out all the stops to get an adequate sample.