Changes

I then used a little clip to attach the Flic button to my pack of cigarettes, which made for an easy protocol: Every time I'd take out a cigarette and smoke, I'd press the Flic button to record the location and timestamp data.  

I then used a little clip to attach the Flic button to my pack of cigarettes, which made for an easy protocol: Every time I'd take out a cigarette and smoke, I'd press the Flic button to record the location and timestamp data.  

Within the first two days it became clear that relying on regular presses of the button would result in recording false-positives: As the button spends a lot of time within the pockets of my trousers, it is prone to accidental short presses or long presses because of other stuff that bumps around in the pocket. Luckily, doing a double press requires a more deliberate action and seemed unlikely to be generated by accident, so I adapted my protocol to only record actual data by doing double clicks.  

Within the first two days it became clear that relying on regular presses of the button would result in recording false-positives: As the button spends a lot of time within the pockets of my trousers, it is prone to accidental short presses or long presses because of other stuff that bumps around in the pocket. Luckily, doing a double press requires a more deliberate action and seemed unlikely to be generated by accident, so I adapted my protocol to only record actual data by doing double clicks.  

I started the data collection on July 13th, 2022 and continued to do the recording protocol until August 16, 2022, giving me a good month worth of baseline data. During this time I did not look at the data beyond spot-checks to make sure the data collection still worked (i.e. no plotting/reasoning about the data during this period).  

I started the data collection on July 13th, 2022 and continued to do the recording protocol until August 16, 2022, giving me a good month worth of baseline data. During this time I did not look at the data beyond spot-checks to make sure the data collection still worked (i.e. no plotting/reasoning about the data during this period).  

The baseline more or less confirmed my initial gut feeling regarding my smoking frequency: I had expected it to be an average of around 10 cigarettes a day, and over the baseline it came out at 8.6 which is slightly lower than guess was.

The baseline more or less confirmed my initial gut feeling regarding my smoking frequency: I had expected it to be an average of around 10 cigarettes a day, and over the baseline it came out at 8.6 which is slightly lower than guess was.