It was difficult to make meaningful comparisons between BI and RPC respondents because the sample sizes were quite different. Because the RPC sample size was relatively small (n=77), differences in pre-test and post-test scores were rarely statistically significant.

The sample size was not large enough to allow further subdivisions by factors such as gender, class level, and librarian.

Response rate influenced sample size for RPC students. Of the students who completed a pre-test prior to their instruction session, 90% of BI students returned the post-test, while only 48% of RPC students returned the post-test. Participation was voluntary, although students were strongly encouraged to return their post-tests. Students receiving BI were given class time to complete the post-test, and professors encouraged their participation. RPC students received post-tests via campus mail.

The BI and RPC groups differed more than we had expected, and this impacted comparability. For example, RPC students (even those without previous instruction) displayed higher levels of knowledge before library instruction than BI students did..