Methodology

The survey instrument was developed by the Publications staff to measure the use of journals and practice characteristics of AAFP active members. The survey was administered to a randomly selected sample of 3,000 AAFP active members. The survey invitations were sent by mail and email. All sample recipients received a mailed invitation, describing the project, a copy of the survey instrument and a link to complete the survey online. The sample recipients were also emailed a short explanation of the project, notification that a copy of the survey was being mailed and a link to complete the survey online. Additional emailed communications were sent to non-responders through the data collection period.

Two hundred twenty-three surveys were returned as a result of the mailing. One hundred ninety-two surveys were completed online. After removing duplicate respondents, a total of 415 usable surveys were used in the final analysis. The number of sample respondents resulted in a margin of error of 4.8% at the 95% confidence interval.

The respondents’ demographic characteristics differed from the AAFP active member population. The respondents were more likely to be older, from rural areas and graduates of US medical school. The respondent data set was weighted to adjust for the differences in the sample.

As in the previous 33 studies, no incentive, prompting or assistance was offered. Any degree of bias, either implied or inherent, is associated solely with the fact that the survey was conducted by the AAFP for its members.