Do Lure Rattles Work? Rattle vs. No Rattle Experiment

I love it that Luke does these tests. I have comments on statistics and acoustics that will help us interpret the results.

Executive Summary
(1) Published audiograms indicate that speckled trout and redfish cannot hear the sound frequencies produced by Battle Rattles (only a few fish species can).
(2) Statistical analysis of Luke’s experimental results do not contradict this view.

Bioacoustics
In addition to dominant frequency produced by the rattles, one needs to consider hearing of the fish. Hearing varies considerably by species (see: https://link.springer.com/content/pdf/10.1007/s11160-012-9297-z.pdf) Audiograms show hearing thresholds, so a high value at a frequency means a high threshold, i.e., worse hearing. Horodsky (2008) studied hearing the family Sciaenidae, including speckled trout & red drum, and found they cannot hear much above 600 Hz (https://journals.biologists.com/jeb/article/211/9/1504/18224/Acoustic-pressure-and-particle-motion-thresholds). In a nutshell, no way are specs or redfish hearing the high-frequency clicks of Battle Rattles. They probably can’t hear any rattles much smaller than a pair of clacking bowling balls. Other fish, however, have higher frequency hearing. Silver perch, for example, probably can hear rattles. Herring have such good hearing at higher frequencies that dolphins hunting herring have to stop echolocating altogether.

Statistics
In these data, we are seeing a small effect size that is well inside the range of likely chance effects. For instance, if we flipped a coin 100 times and got 45 heads and 55 tails, we would not (or should not) argue that our coin proved a meaningful 5% advantage above chance for tails. That result is inside the noise, where chance events likely account for the difference. To see if the observed difference in strike scores is likely due to chance (the null hypothesis) or to rattles (the alternate hypothesis), a simple method is to run a Chi-Square test on the data.
Here’s an online calculator https://www.graphpad.com/quickcalcs/chisquared2/
Enter the two variables [no_rattles, rattles].
Enter the catch results [11, 13]
Enter the expected results, meaning expected by chance [12, 12].
Here’s what the calculator returns:
The two-tailed P value equals 0.6831 
By conventional criteria, this difference is considered to be not statistically significant.
Another way to view this explanation is that a coin flipper making 24 coin tosses would obtain results this extreme or more extreme by chance alone 68% of the time.
If you found a small likelihood of a chance effect, your next question would be whether the effect size is worth the price and bother of the rattles.