I went out tonight from 6:00 to 6:45 and trolled a silver FLUR in the north end of the Loop course. I hooked 3 fish and landed 2, in 45 minutes. Now, I’m not much of a trolling fisherman, but a hookup every 15 minutes looks good for the FLUR.
However, I realized a bias in my experimental design. If I go fishing whenever I feel like it, I may be inclined, unwittingly, to fish the FLURs at more favorable times. Not a good experimental design.
So: I have to randomize the times that I go fishing. The length of days currently (sunrise to sunset) are running about 11 hours. On April 20th the length of day will be around 13.5 hours. The average between the two is a little more than 12 hours. So, I have to randomize the hours after sunrise from sunrise (0) to 12 hours after. There will be a small bias as 12 hours after sunrise is dark now, and on April 20th it will be 1.5 hours before sunset. Aarrgghh
One idea I had was to randomize numbers 0 to6 with a second randomization for after sunrise or before sunset. This will create a slight bias for fishing in the middle of the day, but I think that is better than having the evenings biased.
It never ends. Good thing I’ve just started the study.
Tight Lines!