For the hardcore data heads

[BSK]

Everyone interested has seen all the graphs I post concerning trail-camera data, especially the day-by-day number of bucks picked up on camera, separated by age. I also post graphs of averages per day across years. I often like to look at these numbers over time to see how different "this year" is to previous years. I even break the data out by specific conditions. In essence, buck events on camera per day by just good acorns years, or just acorn failure years. I learn a lot from this stuff.

But while going through years and years of data and comparing them against each other, I began to realize I was making a mistake. The later years looked so much better in bucks photographed per day than the earlier years. Wow, we're really making progress! But I was forgetting a key factor, and that is the number of cameras I'm using. In the early years of my photo censusing, I was generally only running 6-8 cameras per year. However, as of this year, I'm running 20. A big difference exists between the two! A particular date in the early years may have 12 bucks recorded, while this year, the same date 40. What an improvement! Yet, it isn't. 12 bucks recorded from 6 cameras is the exact same level of activity as 40 bucks from 20 cameras. They are both 2.0 bucks per camera.

Because of this "failure to anticipate" changes in data collection, I'm now going back, day-by-day for every year of data collection, and recording the number of cameras running that day, and adjusting the buck numbers to "per camera" instead of just totals. Very eye-opening. It has changed my views on some of my earliest data.

So, for those who are serious data-heads like myself, when comparing year to year data, make sure you are comparing apples to apples. Make sure there hasn't been a change in data collection ability (number of cameras), and if there has been, adjust accordingly (convert your data to "per camera" numbers).

 

[Outdoor Enthusiast]

Interesting. Thanks for sharing.

It will be interesting to see the results from the normalization of the data.

 

[megalomaniac]

Another caveat... not only numbers of cameras, but quality of cameras/ trigger speed/ failure to trigger factors in there. Camera avoidance by shutter clicks, white flash/ red glow in the past skews numbers downward
As well as increased skill in camera placement resulting in increased captures with experience.

I guess what I'm saying is it is impossible to truly compare data from one year to the next when there are so many changing variables with actual sampling....

I DO think the trends over time are accurate, even if sampling is not truly statistically comparable.

My personal philosophy with cameras... Blanket the area to identify all mature bucks using a property. Kill them. Otherwise, the cameras are most useful for fawn recruitment rates which dictate doe harvest for that year.

 

[BSK]

One other caveat to consider Mega. Back when I was only running 6-8 cameras, they were ALL on my best traditional spots, keeping buck numbers high per camera. Now that I'm running 20 cameras, a bunch of those cameras are on "experimental" locations that might turn out to be busts. This decreases the "bucks per camera" number. So, the bucks per camera numbers from the earlier years might be inflated versus later years with more cameras testing more experimental locations.

 

[Andy S.]

All great points, and I could not agree more. Normalizing data collection efforts and date sets is key when extracting key information to glean insights from.

 

[BSK]

Another caveat... not only numbers of cameras, but quality of cameras/ trigger speed/ failure to trigger factors in there. Camera avoidance by shutter clicks, white flash/ red glow in the past skews numbers downward
As well as increased skill in camera placement resulting in increased captures with experience.

I try not to harp on this one too much because there's nothing that can be done about lack of experience (other than give advice to those just beginning a long-term census project). But too often when a new user isn't getting good census results, the reason is often simply lack of experience placing cameras.

 

[Ski]

Back when I was only running 6-8 cameras, they were ALL on my best traditional spots

Assuming you still run cams in the same traditional spots, can you not categorically compare older data against new by focusing only on the locations? The "extra" cams would be a different category altogether.

 

[BSK]

Assuming you still run cams in the same traditional spots, can you not categorically compare older data against new by focusing only on the locations? The "extra" cams would be a different category altogether.

Great idea. Thanks.