Statistical Significance Testing of Sasquatch Sightings and Moon Illumination

*Data used in this blog was current as of October 1, 2024. More data points have been added to the Sasquatch Data Project’s dataset since, and these findings will be updated in the coming months.

Introduction

If sasquatches are just normal animals, wouldn’t we expect them to engage in normal animal behaviors? In today’s blog, we will once again dive into the potential relationship between sasquatch activity and moon illumination, but this time with statistical significance testing. Testing the data for significance allows us to look at relationships within the data and mathematically support the hypothesis that the trends we see are not due to random chance, but due to some other (in this case) biological reason.  This brings up the question, can we prove sasquatches exist with math? The answer is *technically* no, but we can showcase there is data that can mathematically point to them being living, breathing animals. 

Before diving into the data, let’s set the foundation on why moon illumination can be used to learn more about sasquatch behavior. Sasquatches are a species of predator in North America. The predator-prey relationship between scientifically recognized North American animals is highly complex and multifaceted. An avenue of interest for our purposes is the intricate balance of predator and prey activity during extremely low and high moon illumination conditions. It has been found there is an increase in predator activity during both extremely low and high moon illumination (Prugh & Golden, 2013). During high moon illumination, there is greater visibility for both predator and prey. Prey have evolved to recognize this and, in some cases, decrease their activity by 40-70% in an attempt to stay alive (Nersesian, Banks, & McArthur 2011; Taylor et al., 2023). Predators have noted this decrease in prey activity during these times so they typically use this as an opportunity to stake out territory, find mates, and reconvene with their groups, therefore increasing their overall activity.

During periods of extremely low moon illumination, prey have learned predators cannot see as well, so they are more actively foraging and roaming. Predators have learned to capitalize on this opportunity and spend more time hunting under these conditions. Both predator and prey are more active during this time. 

This is a documented phenomenon of known North American predators like wolves, coyotes, bears, and their prey. If sasquatches are to be a typical North American predator, it is expected they would follow this trend of increased activity under extremely low and high moon illumination. We will begin the study by examining witness reports of sasquatches in an attempt to learn more about their behavioral habits and if they correlate to moon illumination in such a manner. 

Method 

At the time of writing, 954 witness testimony reports of sasquatch encounters were parsed from the BFRO (Bigfoot Researchers Field Organization) website and added to the Sasquatch Data Project dataset. Of these 954 reports, 501 have been “Class A” sightings of sasquatches. The classification “Class A” indicates the witness visually observed a large ape of unclassified species. Including only Class A sighting reports allows us to remove doubt and ambiguity that the witness actually encountered a sasquatch and can strengthen our confidence in the results. Out of the 501 Class A sightings, 234 reports included an exact date, allowing for a moon phase and illumination to be found. Moon illumination data in conjunction with the sighting report were obtained from the following states: Alabama, Alaska, Arkansas, Colorado, Delaware, Georgia, Maine, Mississippi, Nebraska, Nevada, New Hampshire, North Carolina, North Dakota, Oregon, Rhode Island, South Dakota, Vermont, and Wyoming.

Moon illumination percentages and their frequencies were then grouped into 5 categories: 0-19%, 20-39%, 40-59%, 60-79%, and 80-100%. All moon illumination groups are inclusive of their boundaries. Extremely low moon illumination, like that under a new moon or close to it, is indicated by the 0-19% category. Extremely high illumination, like that under a full or nearly full moon, is indicated with the 80-100% category.

The frequency data was first plotted into a histogram and a polynomial line of best fit was fitted to the plot (Figure 1). A histogram is a quick way to see the overall pattern of the data. It is a visual way to represent the frequency distribution of the sasquatch sighting tallies under each moon illumination category. The line of best fit produces an R^2 value that tells us how well the line fits the data. The higher the R^2 value, the better the fit. Having an R^2 value of 1 means the trendline fits the data perfectly. This rarely happens in biological sciences, and anything over 0.7 is considered a good fit. Anything over 0.9 is an excellent fit. Assessing the R^2 value is a good way to get an initial read on if there is an overall trend in the data. 

To evaluate whether moon illumination and sasquatch sighting report frequency are correlated, a chi-square test was performed using the frequency counts of sightings within moon illumination groups. A chi-square test is a statistical test that allows us to determine if the trend or pattern seen in categorical data (i.e. moon illumination, hair color, head shape, etc) is due to random chance or an actual physical reason. To perform a chi-square test, you need to establish a null and alternative hypothesis. The null hypothesis, in this case, states sasquatch sighting report frequency does not correlate to moon illumination. This assumes an equal frequency across all moon illumination groups (i.e. no matter what the moon illumination is, we would expect an equal number of sightings to happen in each moon illumination category), and was used as our expected variable during analysis. Our alternative hypothesis states moon illumination does influence the frequency of sasquatch sighting reports. This is where the chi-square test steps in to both test the two hypotheses and investigate if there is a mathematically significant difference between frequency groups. The chi-square test outputs two values: a p-value and a chi-square statistic. The p-value is an important statistical measurement that indicates how likely the frequencies observed could have happened by random chance. In statistics, a p-value of less than or equal to 0.05 is standard and indicates a less than or equal to 5% chance the observed data occurred by random chance. P-values greater than 0.05 indicate the inability to reject the null hypothesis, or the hypothesis that states sasquatch sighting frequency and moon illumination are not correlated. So if we achieve a p-value less than or equal to 0.05, we can be confident there is a biological or physical reason for the trend seen in the sighting frequencies of sasquatches and their corresponding moon illuminations. If our p-value is greater than 0.05, it is very unlikely the observed frequencies are due to a biological or physical reason. We are hoping to achieve a result of less than 0.05 because this would indicate sasquatches behave like normal, biological predators. 

After the chi-square test, the results were further analyzed by comparing the chi-square statistic to the critical value and by using a post hoc pairwise comparison to identify significant differences in frequencies of sightings between different pairings of moon illumination groups. The critical value is a measurement used to again verify that we can reject the null hypothesis. If the chi-square statistic is greater than the critical value, we can once again feel confident the trends and patterns we see in the data are not due to random chance. The post hoc pairwise comparison helps us identify which differences in frequency of  moon illumination categories (i.e. 0-19% vs 40-59%, 20-39% vs 80-100%, etc) are statistically significant. It helps to identify which moon illumination values are producing significantly more reports over others, but backed by math. The post hoc pairwise comparison will produce p-values to help us identify the significant differences. 

A Bonferroni correction was conducted on the p-values from the post hoc test to decrease the risk of a Type I error on the data. A Type I error occurs when a test reveals a false positive result, meaning the p-value was incorrectly under the designated p-value threshold (in this case 0.05). For final analysis, we will look only at the corrected p-values for the pairwise comparison in an effort to reduce the risk of an incorrect result.

Results

Figure 1. A histogram representing the frequency distribution of moon illumination values for sightings. Note: The upper bounds of each group are not inclusive, except in the 80-100% group.

Table 1. Frequency distribution of Class A sasquatch sightings and their corresponding moon illumination groups.

Table 2. Chi-Square test results of the overall frequency distribution of sighting reports and moon illumination groups.

Table 3. Post Hoc Pairwise Comparisons of the different moon illumination groups including Bonferroni correction.

Results

Moon illumination and frequency of sasquatch sighting reports were significantly correlated (R2 = 0.956, Chi-Square statistic: 56.748 > Critical Value: 9.488, p < 0.001). To identify which moon illumination percentages are significant, a post hoc pairwise comparison and Bonferroni correction were performed (Table 3) and the following pairings of moon illumination conditions were found to have a statistically significant difference in the number of reported sightings:

- 0-19% and 40-59% (corrected p-value: 0.002)

- 0-19% and 60-79% (corrected p-value: 0.047)

- 20-39% and 80-100% (corrected p-value: < 0.001)

- 40-59% and 80-100% (corrected p-value: < 0.001)

- 60-79% and 80-100% (corrected p-value: < 0.001)

Two p-values were deemed too large after the Bonferroni correction for moon illumination group comparisons of 0-19% vs 20-38% and 0-19% vs 80-100%. Checking the effect sizes of each pairing, both are relatively small and this supports the finding that they should be removed from the list of statistically significant comparisons, at least for now. 

Discussion

The findings support the hypothesis that sasquatch sighting report frequencies are correlated to extremely low and high moon illumination conditions. The null hypothesis (moon illumination and sighting frequency are not correlated) can be rejected due to the p-value being less than 0.05 and the chi-square statistic being greater than the critical value (Table 2). Sasquatch sighting frequencies during extremely low and high moon illumination conditions points to heightened activity during these times, aligning with scientifically recognized North American predator species behavior. While the increase in sasquatch sightings follows this documented pattern, there could be other factors to influence the increase in apparent activity.

Achieving an overall p-value of <0.05 for the overall analysis and the pairwise comparisons for both extremely high and low moon illumination allows us to reject the null hypothesis. The overall p-value of the chi-square test indicates in itself there is less than a 1% probability these sighting reports occurred in conjunction with their corresponding moon illumination values by random chance. These findings point to one of four scenarios (at least in my mind): 1) If sasquatches are not real animals and all the reports are falsified, the hundreds of sasquatch sighting reporters from the 1920’s to present day, spanning all across the United States, colluded on the dates their reports would have occurred on so this pattern would emerge when someone in the future took the time to look, 2) the less than 1% chance actually happened and these reports just happened to fall into this pattern that matches known animal behavior, 3) there is some sort of human activity increase during these moon illumination conditions therefore leading to more sasquatch sightings, or 4) the findings point to sasquatches being normal animals and taking on normal animal behaviors. Just for fun, let’s explore these options.

Scenario 1 is for all the folks who believe all sasquatch reports are fabricated. If this is true, the hoaxers would have needed to establish this trend in the data either before the BFRO existed or very soon after their inception so they could get this elaborate plan started from the beginning. The BFRO was established in, I believe, 1995. Let’s give the theoretical 234 hoaxers from 18 different states a solid 5 years to come up with a gameplan, and assume they all knew each other in the year 1990 (really this is for ease of calculation). We can easily obtain the populations for each of the states mentioned above along with the number of reports from each state that have a moon illumination percent associated with it (the sample value). We can then use the population and sample sizes to get a very rough probability calculation that all of these people even knew each other. Just to give the faux hoaxers even more of a leg up in this scenario, we will have a loose definition of “knowing each other” and include a friend of a friend relationship to one another, creating more of a social chain over a social circle. Using a simple, very rough population calculation, the chances these 234 people were somehow connected in any way in the year 1990 is approximately 0.44%. This is actually a higher percentage than I was expecting, but is still far less than a 1% chance. This is assuming every single person followed through with the scheme as well, and that factor is not included in the calculation. This calculation also does not include a 1990 social dynamic factor, and an accurate estimate on the level of connection between states (in this case we have an overestimate of the state-to-state connection). So this result of 0.44% is actually quite generous. In an effort to get a more accurate probability, I also conducted a Monte Carlo simulation and included factors like a local community factor (the chances people know each other in the same state), an interstate factor (the chance people in different states know each other), a technology factor (limited long-distance technological communication in 1990), a profession factor (knowing someone because of a job), and a family factor (extended family across state lines). A Monte Carlo simulation basically runs a set number of simulations on the scenario in question and produces a probability for each simulation. It then averages the probabilities to give an overall probability that the scenario could happen. In this case, the simulation was attempting to make the 234 person connection with the given sample sizes, populations, and social factors. The result was a resounding 0.000000. The simulations were never able to achieve the result within 10,000 separate simulations, therefore giving us a basically impossible scenario. Just for fun, I ran the simulation 50,000 times and still achieved the result of 0. So, after looking at the calculation and simulation results, somewhere between 0 and 0.44%, but more so leaning towards 0, is the chance this is an elaborate hoax pulled off by a bunch of very bored people in the 1990s. We can feel confident this is not the result of hoaxing, let’s move on to scenario 2: A freak of nature occurred and these results are due to random chance.

Typically in academic writing, a p-value less than 0.001 is just reported as so for ease of interpretation. In this case, though, I would like to mention the exact p-value calculated for the overall trend in sasquatch sighting frequency and moon illumination was 1.397x10^-11. In plain text that is 0.00000000001397. So there is a 0.000000001397% chance the results seen in the data happened by random chance. This would be nearly impossible, and there isn’t much more to say about it. We can confidently rule this out as a possibility. 

A third reason for the increase in sightings could be a human factor.The working hypothesis from sasquatch researchers is that under full moon conditions humans can see better therefore leading to more sasquatch sightings. From the research that has been conducted on the luminance of a new vs full moons, I do not think this is the case. Luminance during a new moon versus that of a full moon only differs by one order of magnitude (0.03 & 0.26 lux, respectively) and is not enough to make a significant difference in a human’s ability to see at night (Todd et al. 2015). Even under full moon conditions, the luminance produced by the moon only is less than what is recommended for common human tasks. For example, the recommended luminance for a movie theater’s house lighting during the film is 1-2 lux and does not include the luminance brought on by the film itself (Rea, 2000). An illumination of 30 lux is recommended to successfully navigate a public space or read (Rea, 2000). Therefore, the terrestrial luminance of the moon is less than sufficient in aiding human sight (Todd et al. 2015). Additionally, this study did not take into consideration time of day and looked at the influence of the moon’s illumination and phase as a whole. Of the 234 data points used in this blog, 103 occurred during the daytime which I defined as 7AM to 7PM. Obviously, these times fluctuate through the year and are based on location so there will be some discrepancy in an exact number. For now though, nearly half of the data that produced these results happened during the day, so there is some kind of factor outside of the illumination itself influencing sasquatch behavior.

With a human’s sight out of the picture, perhaps humans are more active in general under extremely low and high moon illumination, leading to an increase in sasquatch sightings. Humans are predators after all, so wouldn’t we expect them to also follow this trend of increased activity during extremely low and high moon illumination conditions? Not quite. The main difference between the predator behaviors of humans and sasquatches is that humans are not nocturnal, while it is suspected sasquatches are. Nocturnal predators are far more influenced by the lunar cycle compared to their diurnal counterparts. Humans following the lunar cycles is driven more so by cultural norms, personal preference, and sleep patterns. It has been found under full moons humans sleep less and take longer to fall asleep (Casiraghi et al. 2021). It was found that throughout the duration of the lunar cycle, the difference in the amount of sleep participants got overall differed from 20-90 minutes depending on the lunar phase. The study found under new moon conditions humans slept up to 25% more than under full moon conditions. It is worth noting the researchers were also studying how artificial light interfered with sleep cycles. So possibly if humans are awake longer, at least under full moons, they would have more of an opportunity to see a sasquatch. This could possibly be a reason for the increase in high moon illumination sightings, but it does not explain the increase in sightings during new moons.

The last scenario at this point seems the most likely. Sasquatches are normal animals doing normal animal things. If we had gotten these results for, say, a coyote, no one would bat an eye. A sigh & exasperated “DUH” would erupt from the scientific community. But for some reason an ape that stays out of everyone’s way and is just living their best predator life makes everyone uncomfortable. *Shrug* Proceeding with the most logical of the scenarios, let’s look at the data.

This study found sightings of sasquatches are most frequently reported under extremely high moon illumination values (80-100%) and this increase is significant compared to the middle moon illumination values, indicating sasquatches are more mobile during these periods of time. While the full moon’s illumination was not enough to make a difference in human sight, it is enough to cause a noticeable difference in visibility for nocturnal animals (Auselbrook et al. 2022). It is thought sasquatches are nocturnal due to a number of reports mentioning the presence of eye shine. The Sasquatch Data Project’s dataset has currently noted 37 Class A sighting reports of sasquatches that claimed the presence of eyeshine. There have been 7 reports that did not claim eyeshine in a situation that “should” have produced it (i.e. headlights directly on the eyes, a flashlight shining on the face, etc.). If sasquatches are truly nocturnal (though I am not quite convinced, but that is for a later blog) changes in the moon illumination even by just an order of magnitude could potentially have a significant impact on their ability to see, therefore increasing their mobility and activity levels as seen in other nocturnal animal species (Auselbrook et al. 2022). This increase in visibility for nocturnal predators would allow them to cover more ground in staking out territory, find family groups and mates easier, and help them hunt a rogue prey animal who didn’t get the memo about being eaten easier. Chimpanzees, though diurnal like all other ape species, have adopted nocturnal behaviors due to human intrusion into their environment. One group in particular was found to be more actively foraging and on the hunt for mates during nights with a full moon (Krief et al. 2014). So it is possible sasquatches could be partaking in similar activities during these conditions since it is known ape behavior. To truly understand the extent of moon illumination on sasquatch activity, we need to study how sighting frequency increases and decreases with moon illumination during day and night instead of as a whole. At this time more data needs to be collected to conduct that study, but preliminary analysis shows regardless of time of day the results of this blog stay consistent. So, for now a possible explanation for the increased sighting report frequency under high moon illumination could be that sasquatches can see significantly better compared to new moon conditions, and are more active. 

Increased sightings under new moon conditions were also statistically significant, and can also support the hypothesis they are more active during these conditions. This could potentially be a fruitful time for them to hunt due to the predator-prey relationship. As stated previously, prey are more active under lower moon illumination, therefore increasing the activity levels of predators. Sasquatches are likely following this trend. They may also be more active under low moon illuminations due to decreased visibility, so it may take them longer to forage and hunt or meet up with others in their family groups. In the same chimpanzee study mentioned previously, a different group of chimpanzees were actually more active under new moon conditions as opposed to a full moon night (Krief et al. 2014). They were partaking in the majority of their hunting and foraging during these conditions potentially to avoid humans or due to the maturation stage of the particular crop they were interested in (Kreif et al. 2014). Potentially sasquatches could be exhibiting similar behaviors under low moon illumination. They could be attempting to avoid detection from humans, so they wait for lower moon illumination to come in closer to human proximity thinking it will help keep them under the cover of darkness. It is difficult to say at this point what else may be driving this behavior since sasquatches, if truly nocturnal and not showing cathemerality, would be the only nocturnal ape species. We can look at the behaviors of the few nocturnal primates in existence (like the loris and lemur), but they are not top predators in their environments so their motives may be a bit different from a sasquatches. Also, while it is fascinating how chimpanzees have adopted nocturnal behaviors, it is important to once again remember they are not truly nocturnal, so there will be differences in the behaviors of sasquatches and chimpanzees. Looking only at apex predators in North America who exhibit similar behavioral patterns, we can say with some confidence this trend is related to obtaining a food source in some capacity. 

Conclusion

I am quite excited by these results, but it is important to mention that I still have thousands of reports to parse off the BFRO website. While I have more than enough data points to support a chi-square test (only 30 data points are needed to run analysis for this test. I used 234), more data is always better to strengthen the results. I am eager to see how these findings may change over time, but for now there does seem to be a strong relationship between moon illumination and sasquatch sighting reports. To me, this supports the idea that sasquatches are just normal apes doing normal ape things and follow known predator behavior.

*Updated 11/1/24