Thesis Meeting

I was getting frustrated with the data I’ve been fitting, so it was good to finally meet with them to talk about it. The goal for the past month was to take a previously published data set to compare the results I get from the automated pipeline to the published results. We, of course, were hoping to find that the pipeline was getting comparable results, because that would bolster my pipeline.

Instead, I found that my data showed the exact opposite trend as the published data. I did everything I could think of to re-analyze and re-fit the data to see if I could extract their trends. In the end, my data just doesn’t show the same trend. This is a potentially interesting and very exciting result. One difference between the way I’m analyzing the data and they way they are is that all the points in my spectra are temporally adjacent, whereas they extract all points in the observation according to flux level with no attempt to keep temporally adjacent points together. What we’re thinking this might do is smear out the time scales of change in some of the parameters. This may produce spurious relationships between parameters that can’t really show the true picture of variability in the source.

I’ll be testing this theory by re-extracting my data using their technique. If I find their same trends, that will be a strong case that their work was not the right way to do things. If I find something else, then we’ll have to look more closely at my work to see what’s going on.