Skip to contents

Describe each person's series

Source: R/stats_describe.R
describe_persons.Rd

One row per person per variable: how much data they contributed, where they sit, how much they move, and how strongly each occasion carries into the next.

Usage

describe_persons(
  data,
  id,
  vars = NULL,
  time = NULL,
  subject = NULL,
  detail = c("basic", "full"),
  pac_quantile = 0.9,
  pac_cutoff = NULL,
  pac_direction = c("absolute", "increase", "decrease"),
  variable = NULL,
  sort_by = NULL,
  decreasing = FALSE,
  n = NULL
)

Arguments

data

Data frame of repeated measures.

id

Person/unit ID column.

vars

Columns to describe (any fit_lm() selector). Defaults to every numeric column other than id and time.

time

Optional ordering column. Supply it: without it, "successive" means the order the rows happen to be in.

subject

Optional person(s) to describe. Defaults to everyone.

detail

"basic" (the default) or "full", which adds distribution shape, floor/ceiling occupancy, the probability of acute change, the person's linear trend, and the longest run of identical consecutive values.

pac_quantile

Quantile of the pooled successive changes that defines an "acute" one. The convention is 0.9.

pac_cutoff

Use this change size instead of a quantile of the data. Supply it to reproduce a published cutoff, or to make pac comparable across datasets.

pac_direction

Whether an acute change means any large change ("absolute"), or specifically a large rise or fall.

variable

Optional variable(s) to keep.

sort_by

Optional column of the result to sort by, e.g. "rmssd".

decreasing

Sort order when sort_by is supplied.

n

Optional number of rows to keep.

Value

A data.frame of one row per person per variable, with class idiographic_descriptives.

Details

n, missing

Usable and missing occasions – the compliance question, asked per person rather than for the sample as a whole.

mean, sd

Level and overall dispersion.

rmssd

Root mean square successive difference: the average occasion-to-occasion change. This is a different quantity from sd, and the pair is more informative than either alone – a slow drift gives a large sd with a small rmssd, and rapid oscillation gives the reverse.

autocor

Lag-1 autocorrelation, the usual measure of inertia or carry-over: how much of where a person is now is explained by where they just were.

pac

Probability of acute change (detail = "full"): how often this person's occasion-to-occasion change clears a bar set by the whole sample – by default the 90th percentile of everyone's changes. rmssd says how large the changes are and is dominated by a few big swings; pac says how often a large one happens and is not. Jahng, Wood and Trull (2008) recommend the pair together. Being sample-relative, pac is not comparable across datasets unless pac_cutoff is supplied.

span, gap_median, gap_max

Only when time is given. How long the person was observed and how far apart their occasions were. Worth reading before trusting any lag: gap_max far above gap_median means "the previous occasion" is not a constant amount of time.

Successive differences and the autocorrelation are computed on adjacent occasions in time order within each person, using only pairs where both values are present. They are never taken across a person boundary.

Examples

describe_persons(srl, "name", time = "day", vars = c("effort", "efficacy"))
#> PERSON DESCRIPTIVES
#>   Grouping    name
#>   Time        day
#>   People      36
#>   Variables   2
#> 
#>   subject   variable     n   miss     mean   median       sd      min       max    rmssd   autocor      span   gap_median   gap_max
#> -----------------------------------------------------------------------------------------------------------------------------------
#>   Aisha     efficacy   156      0   56.919   61.765   21.034    0.000   100.000   26.881     0.169   155.000        1.000     1.000
#>   Alice     efficacy   156      0   35.761   36.364   20.614    0.000   100.000   28.790     0.024   155.000        1.000     1.000
#>   Anika     efficacy   156      0   50.099   47.692   19.373    0.000   100.000   28.656    -0.124   155.000        1.000     1.000
#>   Astrid    efficacy   156      0   72.401   78.378   20.078   10.811   100.000   28.496    -0.008   155.000        1.000     1.000
#>   Bjorn     efficacy   156      0   52.707   51.852   21.803    3.704   100.000   30.443     0.023   155.000        1.000     1.000
#>   Bob       efficacy   154      2   78.219   77.108   16.717    0.000   100.000   23.209     0.028   155.000        1.000     1.000
#>   Charlie   efficacy   156      0   46.239   46.667   17.767    0.000   100.000   25.857    -0.061   155.000        1.000     1.000
#>   Diana     efficacy   156      0   86.699   90.000   14.696    0.000   100.000   21.459    -0.066   155.000        1.000     1.000
#>   Erik      efficacy   156      0   28.050   27.586   27.530    0.000   100.000   39.069    -0.014   155.000        1.000     1.000
#>   Eve       efficacy   156      0   49.222   52.459   31.279    0.000   100.000   45.673    -0.076   155.000        1.000     1.000
#>   Fatima    efficacy   154      2   67.418   70.588   17.420    0.000   100.000   26.882    -0.187   155.000        1.000     1.000
#>   Frank     efficacy   156      0   61.311   61.458   15.797   16.667   100.000   23.334    -0.095   155.000        1.000     1.000
#> 
#>   ... 60 more rows.
#> 
#>   sd    = overall spread;  rmssd = occasion-to-occasion change
#>   autocor = lag-1 carry-over (inertia)

# One person, or the most volatile few -- without reaching into the result.
describe_persons(srl, "name", time = "day", subject = "Aisha")
#> PERSON DESCRIPTIVES
#>   Grouping    name
#>   Time        day
#>   People      1
#>   Variables   9
#> 
#>   subject   variable       n   miss     mean   median       sd     min       max    rmssd   autocor      span   gap_median   gap_max
#> ------------------------------------------------------------------------------------------------------------------------------------
#>   Aisha     control      156      0   52.389   51.515   25.507   0.000   100.000   35.298     0.031   155.000        1.000     1.000
#>   Aisha     efficacy     156      0   56.919   61.765   21.034   0.000   100.000   26.881     0.169   155.000        1.000     1.000
#>   Aisha     effort       156      0   77.646   82.540   18.080   0.000   100.000   23.724     0.132   155.000        1.000     1.000
#>   Aisha     help         156      0   48.244   49.000   18.839   0.000    88.000   28.131    -0.124   155.000        1.000     1.000
#>   Aisha     monitoring   156      0   25.287   18.421   24.259   0.000   100.000   33.391     0.052   155.000        1.000     1.000
#>   Aisha     organizing   156      0   61.270   69.355   27.941   0.000   100.000   38.844     0.003   155.000        1.000     1.000
#>   Aisha     planning     156      0   60.795   63.636   21.219   0.000    95.455   29.409     0.012   155.000        1.000     1.000
#>   Aisha     social       156      0   55.753   56.250   20.827   0.000   100.000   27.481     0.124   155.000        1.000     1.000
#>   Aisha     value        156      0   60.986   63.889   20.730   5.556   100.000   26.280     0.194   155.000        1.000     1.000
#> 
#>   sd    = overall spread;  rmssd = occasion-to-occasion change
#>   autocor = lag-1 carry-over (inertia)
describe_persons(srl, "name", time = "day", variable = "effort",
                 sort_by = "rmssd", decreasing = TRUE, n = 5)
#> PERSON DESCRIPTIVES
#>   Grouping    name
#>   Time        day
#>   People      5
#>   Variables   1
#> 
#>   subject   variable     n   miss     mean   median       sd     min       max    rmssd   autocor      span   gap_median   gap_max
#> ----------------------------------------------------------------------------------------------------------------------------------
#>   Karin     effort     156      0   61.176   73.913   34.061   0.000   100.000   48.061     0.001   155.000        1.000     1.000
#>   Erik      effort     156      0   42.330   47.368   35.159   0.000   100.000   46.286     0.126   155.000        1.000     1.000
#>   Lars      effort     156      0   50.701   53.488   31.961   0.000   100.000   43.756     0.059   155.000        1.000     1.000
#>   Frank     effort     156      0   44.409   37.037   28.913   0.000   100.000   43.589    -0.140   155.000        1.000     1.000
#>   Hiroshi   effort     156      0   23.449   10.000   26.630   0.000   100.000   38.260    -0.066   155.000        1.000     1.000
#> 
#>   sd    = overall spread;  rmssd = occasion-to-occasion change
#>   autocor = lag-1 carry-over (inertia)