Collect and Process Physical Activity Direct Observation Data

Description

This provides a free and easy way to document and annotate physical activity behaviors using direct observation.

Core functions

data_collection_program

compendium_reference

Associated References

Hibbing PR, Ellingson LD, Dixon PM, & Welk GJ (2018). Adapted Sojourn Models to Estimate Activity Intensity in Youth: A Suite of Tools. Medicine and Science in Sports and Exercise. 50(4), 846-854. doi:10.1249/MSS.0000000000001486.

Note

'Observation' functions accept further arguments that are passed to functions from the 'svDialogs' package. Doing so may improve your experience using the 'Observation' package. See the package vignette for more information.

Examples

# Example of non-interactive use
data(example_data)
compendium_reference(example_data, FALSE, kids = "yes", mvpa = "yes")

# Example of interactive use
if (interactive()) {
  observation_data <- data_collection_program()
  full_data <- compendium_reference(observation_data)
}

Verify Correctness of User-Specified Interactive Preference

Description

Verify Correctness of User-Specified Interactive Preference

Usage

check_interactive(interactive)

Arguments

Consult the Compendium of Physical Activities to Encode Direct Observation Intensities

Description

Consult the Compendium of Physical Activities to Encode Direct Observation Intensities

Usage

compendium_reference(obs_data, interactive = TRUE, ...)

Arguments

Value

A data frame fully annotated with intensity values

Note

If 'interactive = TRUE', but R is not being used interactively, a message is issued and an override is implemented to set 'interactive = FALSE'.

See Also

Other processing functions: interactive_compendium_reference(), manual_compendium_reference()

Examples

data(example_data)

compendium_reference(example_data, FALSE, kids = "yes", mvpa = "yes")
if (interactive()) {
    compendium_reference(example_data)
}

Collect Direct Observation Data

Description

A generic-like function for collecting direct observation data manually or interactively, depending on user-specified preference and session capabilities.

Usage

data_collection_program(interactive = TRUE, ...)

Arguments

Note

If 'interactive = TRUE', but R is not being used interactively, a message is issued and an override is implemented to set 'interactive = FALSE'.

See Also

example_data

Other collection functions: interactive_data_collection_program(), manual_data_collection_program()

Examples

if (interactive()) {
data_collection_program()
}

# Load Sample of interactively-collected data for comparison
#    with manually-entered data

data("example_data")

# Example of manually defining input variables -------------------------

id <- "Test_ID"

timestamps <- c("2018-05-06 02:40:37", "2018-05-06 02:40:46",
    "2018-05-06 02:40:59", "2018-05-06 02:41:14",
    "2018-05-06 02:41:27", "2018-05-06 02:41:37",
    "2018-05-06 02:41:48", "2018-05-06 02:42:03",
    "2018-05-06 02:42:24", "2018-05-06 02:42:39",
    "2018-05-06 02:42:53")

activities <- c("sitting still", "sitting stretching",
    "sitting shoulder press", "sitting cycling",
    "standing still", "standing stretching",
    "standing shoulder press", "walking slowly",
    "walking quickly", "jumping jacks")

durations <- c(9, 13, 15, 13, 10, 11, 15, 21, 15, 14.460825920105)

Tree_Intensity <- c("Sedentary", "Sedentary/Light",
    "Light/Moderate", "Indeterminate", "Light", "Light",
    "Light/Moderate", "Light", "MVPA", "Light/Moderate")

seated <- c("yes", "yes", "yes", "yes", "no", "no", "no",
    "no", "no", "no")

large_muscles_moving <- c("no", "yes", "yes", "yes",
    "no", "yes", "yes", "yes", "yes", "yes")

slow <- c(NA, "yes", "yes", "no", NA, "yes", "yes",
    "no", "no", "no")

slowed_by_resistance <- c(NA, "no", "yes", NA, NA,
    "no", "yes", NA, NA, NA)

ambulation <- c(NA, NA, NA, NA, NA, NA, NA,
    "yes", "yes", "no")

light_walking <- c(NA, NA, NA, NA, NA, NA,
    NA, "yes", "no", NA)

# Example of using the manual program ----------------------------------

manual_data <- data_collection_program(
  interactive = FALSE, id = id, timestamps = timestamps,
  activities = activities, durations = durations, seated = seated,
  large_muscles_moving = large_muscles_moving, slow = slow,
  slowed_by_resistance = slowed_by_resistance, ambulation = ambulation,
  light_walking = light_walking
)

# Comparing output of interactive vs manual program --------------------

test_names <- intersect(names(example_data), names(manual_data))
test_names <- setdiff(names(test_names), "duration_s")

all.equal(
  example_data[ ,test_names],
  manual_data[ ,test_names],
  scale = 1,
  tolerance = 5
)

Sample data collected interactively with direct observation program.

Description

A data set assembled by running the interactive direct observation program and entering a variety of activities.

Usage

example_data

Format

A data frame with 10 rows and 15 variables:

id

The user-inputted identification code for the trial

User_Timestamp

The user-inputted start time for the activity

Auto_Timestamp

The automatic timestamp generated by button click, retained as a backup if an error occurs in User_Timestamp

dayofyear

The numeric day of the year (i.e., julian date)

minofday

The numeric minute of the day (e.g. 0 for midnight and 1439 for 23:59)

Activity

The user-inputted activity description

duration_s

The duration of the activity (in seconds), based on User_Timestamp

auto_duration_s

The duration of the activity (in seconds), based on Auto_Timestamp

Tree_Intensity

The pre-classified activity intensity, based on user input to the program's decision tree

seated

User-inputted response to the seated prompt

large_muscles_moving

User-inputted response to the large-muscles-moving prompt

slow

User-inputted response to the slow prompt

slowed_by_resistance

User-inputted response to the slowed-by-resistance prompt

ambulation

User-inputted response to the ambulation prompt

light_walking

User-inputted response to the light-walking prompt

Helper Function for Interactive Intensity Coding Process.

Description

Interface for looking up Compendium values to assign an intensity to an activity.

Usage

interactive_comp_lookup(incorrect_entries, Levels, compendium, ...)

Arguments

Interactively Consult the Compendium of Physical Activities to Encode Direct Observation Intensities

Description

Interactively Consult the Compendium of Physical Activities to Encode Direct Observation Intensities

Usage

interactive_compendium_reference(obs_data, ...)

Arguments

Value

A data frame fully annotated with intensity values

Note

interactive_compendium_reference is called within compendium_reference when interactive = TRUE and the session is interactive.

See Also

Other processing functions: compendium_reference(), manual_compendium_reference()

Examples

if (interactive()) {
data(example_data)
compendium_reference(example_data)

observation_data <- data_collection_program()
compendium_reference(observation_data)
}

Interactively Collect Direct Observation Data

Description

Interactively Collect Direct Observation Data

Usage

interactive_data_collection_program(...)

Arguments

Value

A data frame of observation data

Note

interactive_data_collection_program is called within data_collection_program when interactive = TRUE and the session is interactive.

This program relies heavily on dialog boxes generated by the 'svDialogs' package. Default measures are in place to ensure baseline functionality across operating systems and R interfaces. However, a better experience may be achievable by passing additional arguments to the dialog functions via '...', as the vignette for 'Observation' describes.

See Also

[example_data]

Other collection functions: data_collection_program(), manual_data_collection_program()

Examples

if (interactive()) {
data_collection_program()
}

Interactively Pre-Classify Activity Intensity

Description

Interactively implement the pre-classification decision tree described at the end of Supplemental Document 3 from Hibbing et al. (2018, *Med Sci Sports Exerc*).

Usage

interactive_tree_intensity(...)

Arguments

Manually Encode Direct Observation Intensities by Cross-Referencing the Compendium of Physical Activities

Description

Manually Encode Direct Observation Intensities by Cross-Referencing the Compendium of Physical Activities

Usage

manual_compendium_reference(
  obs_data,
  kids = c("yes", "no"),
  mvpa = c("yes", "no")
)

Arguments

Value

A list with three elements. The first (data) is the original data, where a Final_Intensity column has been populated to the extent possible, with "Indeterminate" listed for the other activities. The second element (indeterminate) is another list, which has one element for each activity in data with Final_Intensity = "Indeterminate". The elements of indeterminate are named after the activity they represent, and the contents are a subset of the Compendium giving suggested matches for the activity, based on its description in data. The third element (compendium) gives the entire compendium, which can be manually cross-referenced for cases where the suggested matches in indeterminate do not give a suitable match.

See Also

Other processing functions: compendium_reference(), interactive_compendium_reference()

Examples

data(example_data)

example_data_processed <- compendium_reference(example_data, FALSE,
   kids = "yes", mvpa = "yes")

if (interactive()) {
 View(example_data_processed$data)
 View(example_data_processed$indeterminate[[1]])
 View(example_data_processed$compendium)
}

Manually Collect Direct Observation Data

Description

Manually Collect Direct Observation Data

Usage

manual_data_collection_program(
  id,
  timestamps,
  activities,
  durations = NULL,
  seated = NA,
  large_muscles_moving = NA,
  slow = NA,
  slowed_by_resistance = NA,
  ambulation = NA,
  light_walking = NA
)

Arguments

Value

A dataframe of pre-processed direct observation data generated from minimal input

Note

By default, activity duration is automatically calculated, which is accomplished via diff.POSIXt and thus requires n + 1 timestamps, where n is the number of activities. Alternatively, durations can be manually specified via the durations parameter, in which case only n timestamps are needed, corresponding to the onset of each activity.

See Also

Other collection functions: data_collection_program(), interactive_data_collection_program()

Manually Pre-Classify Activity Intensity

Description

Manually implement the pre-classification decision tree described at the end of Supplemental Document 3 from Hibbing et al. (2018, *Med Sci Sports Exerc*).

Usage

manual_tree_intensity(prompt_responses)

Arguments

Note

The vector of responses must match the structure indicated in the example, i.e., a named vector answering prompts in the following order: participant seated; large muscles contracting; slow contractions; contractions slowed by resistance; activity is ambulatory; ambulation is slow enough to be considered light activity rather than moderate or vigorous.

Examples

prompt_responses <- structure(c("yes", "no", NA, NA, NA, NA),
    .Dim = c(6L, 1L), .Dimnames = list(c("seated",
      "large_muscles_moving", "slow", "slowed_by_resistance",
      "ambulation", "light_walking"),
    "1"))

Observation:::manual_tree_intensity(prompt_responses)