Skip to contents

5. Accounting for Study Protocol

Source: vignettes/chapter5_StudyProtocol.Rmd
chapter5_StudyProtocol.Rmd

When you use an accelerometer in a study, you are likely to give your study participants specific instructions on when they should start wearing the accelerometer, for how many days, and whether or not they are expected to take the accelerometer off during specific activity types or parts of the day. Further, it may be that you turned on the accelerometers hours or even days before you gave it to the participant or stopped it hours or days after you received it back. Your knowledge about all these aspects of your study protocol can be used by GGIR to mask certain periods of time in the recording. This is important because this information is not necessarily obvious from the recorded data. For instance, when a recording is started and dispatched to the participant via mail, the time during which the devices are in transit and not worn may be impossible to distinguish from a participant wearing the accelerometer and commuting.

Selecting/Masking the data

It is important that GGIR analyses masks all data outside the time window for which the participant was instructed to wear the accelererometer. Study protocols differ in duration and expected wear period, which is why GGIR offers a variety of ways to account for study protocol.

The main parameter to do this is data_masking_strategy, and it requires a numeric value indicating one of the following strategies:

  • data_masking_strategy = 1 to indicate the a specific number of hours should be masked from the start and/or the end of the recording, specified with parameters hrs.del.start and hrs.del.end, respectively.

  • data_masking_strategy = 2 to indicate that only the data between the first and the last midnight in the recording should be considered.

  • data_masking_strategy = 3 to indicate that only the most active X 24-h blocks starting any time in the day should be used, where X is specified by parameter ndayswindow. Note this can be combined with beforementioned parameters hrs.del.start and hrs.del.end, which will trim this window at the start and end.

  • data_masking_strategy = 4 to indicate that only the data after the first midnight should be considered.

  • data_masking_strategy = 5 is similar to data_masking_strategy = 3, yet it selects X complete calendar days where X is specified by parameter ndayswindow.

Additionally, the maximum duration after the start of the recording that the accelerometer is expected to be worn can also be set with parameter maxdur for number of 24 hour blocks or parameter max_calendar_days for number of calendar days.

Key parameters

  • data_masking_strategy

  • hrs.del.start

  • hrs.del.end

  • ndayswindow

  • maxdur

  • max_calendar_days

(Part of) variable name Description Report(s)
data exclusion strategy A log of the decision made when calling g.impute: value=1 mean ignore specific hours; value=2 mean ignore all data before the first midnight and after the last midnight part2_summary.csv
n hours ignored at start of meas Number of hours ignored at the end of the measurement (if data_masking_strategy = 1) or at the end of the ndayswindow (if data_masking_strategy = 3 or 5). A log of decision made in part2.R part2_summary.csv
n hours ignored at end of meas Number of hours ignored at the start of the measurement (if data_masking_strategy = 1) or at the start of the ndayswindow (if data_masking_strategy = 3 or 5) A log of decision made in part2.R part2_summary.csv
n days of measurement after which all data is ignored Number of days of measurement after which all data is ignored (if data_masking_strategy = 1, 3 or 5) A log of decision made in part2.R part2_summary.csv