Today was the big day! I’d been planning it for months. I was going to meet in the classroom of the cohort that I’ve been following for 3 years, and bribe them into taking my last survey by offering pizza!
But instead, I got COVID. Talk about bad timing… Thankfully, my loyal supervisors Torstein and Tove went in my place, and survey results are now ticking in. 🙂
There aren’t many longitudinal studies in information literacy, so this aspect of my research is one of the things that makes it unique. Few researchers have the opportunity to conduct long-term projects like this one, but since I had 4 years to complete my PhD, I had the chance.
The pandemic, however, has taken its toll on the cohort that I’ve been following. They started just half a year before the lock-down, at which point their classes went online for nearly 2 years. For many students, this made it (understandably) less appealing to continue their college education. At the start of their Bachelor in psychology in the fall of 2019, there were 98 students. Now, at the end of the program, there are only 32 left. To make matters worse, only 14 of those 32 are from the original cohort. That’s an 86% attrition rate! (The other 18 that are finishing their Bachelor degrees now have most likely transferred from other programs or universities.)
The small sample size represents a challenge for my research design and data analyses as well. I’d planned on following individual students over 3 years with a repeated measures (within-group) design. However, since there were only 14 of students remaining in the end, the sample size would be quite small. But luckily, with repeated measures, statistical inference can be made with fewer participants. Alternatively, I could analyze data from the original 98 and the final 32 participants with an independent measures (between-group) design. With an independent measures design, more participants are necessary in order to make statistical inference than with a repeated measures design. So, once I see how many students have responded to the survey, I’ll weigh the pro’s and con’s of both designs before making the decision of which to use in the statistical analyses.
So, as we say in Norwegian, det var dagens hjertesukk – “that was today’s heart sigh” (today’s worries). I’m thankful in any case for the data that we have managed to collect (and for that my COVID symptoms are negligible!).