Last chance! 50% off unlimited learning
Sale ends in
r lifecycle::badge('experimental')
This function implements ITSA method described in the paper 'Conducting interrupted time-series analysis for single- and multiple-group comparisons', Ariel Linden, The Stata Journal (2015), 15, Number 2, pp. 480-500
This function further requires the installation of 'sandwich' and 'lmtest' in
order to work. These packages can be installed from CRAN using
install.packages().
create_itsa(
data,
metrics = NULL,
before_start = NULL,
before_end = NULL,
after_start = NULL,
after_end = NULL,
ac_lags_max = 7,
return = "table"
)When 'data' is passed to return, a data frame with the following columns:
metric_name: Name of the metric being analyzed.
beta_2: Coefficient for the immediate treatment effect.
beta_3: Coefficient for the treatment effect over time.
beta_2_pvalue: P-value for the immediate treatment effect.
beta_3_pvalue: P-value for the treatment effect over time.
AR_flag: Logical flag indicating whether autocorrelation was detected.
error_warning: Error or warning message if applicable.
Person Query as a dataframe including date column named
MetricDate. This function assumes the data format is %Y-%m-%d as is
standard in a Viva Insights query output.
A character vector containing the variable names to perform the interrupted time series analysis for.
String specifying the start date of the 'before'
time period in %Y-%m-%d format. The 'before' time period refers to the
period before the intervention (e.g. training program, re-org, shift to
remote work) occurs and bounded by before_start and before_end
parameters. Longer period increases likelihood of achieving more
statistically significant results. Defaults to earliest date in dataset. If
not provided, this defaults to the earliest date in the dataset.
String specifying the end date of 'before' time
period in %Y-%m-%d format. If NULL, an error will be raised, as this
value is required.
String specifying the start date of the 'after'
time period in %Y-%m-%d format. If NULL, this will default to the value
of before_end. The 'after' time period refers to the period after the
intervention occurs and is bounded by after_start and after_end
parameters. Longer periods increase the likelihood of achieving more
statistically significant results.
String specifying the end date of the 'after' time
period in %Y-%m-%d format. Defaults to the latest date in the dataset.
Numeric value specifying the maximum lag for the autocorrelation test. The Ljung-Box test is used to check for autocorrelation in the model residuals up to this specified number of lags. Higher values check for longer-term dependencies in the time series data.
String specifying what output to return. Defaults to "table". Valid return options include:
'plot': return a list of plots.
'table': return data.frame with estimated models' coefficients and
their corresponding p-values You should look for significant p-values in
beta_2 to indicate an immediate treatment effect, and/or in beta_3 to
indicate a treatment effect over time
Aleksey Ashikhmin alashi@microsoft.com
This function uses the additional package dependencies 'sandwich' and 'lmtest'. Please install these separately from CRAN prior to running the function.
As of May 2022, the 'portes' package was archived from CRAN. The dependency
has since been removed and dependent functions Ljungbox() incorporated into
the wpa package.
if (FALSE) {
# Returns summary table
create_itsa(
data = pq_data,
metrics = c("Collaboration_span", "Internal_network_size"),
before_end = "2024-07-01",
after_start = "2024-07-01",
ac_lags_max = 7,
return = "table"
)
# Returns list of plots
plot_list <-
create_itsa(
data = pq_data,
metrics = c("Collaboration_span", "Internal_network_size"),
before_end = "2024-07-01",
after_start = "2024-07-01",
ac_lags_max = 7,
return = "plot"
)
# Extract a plot as an example
plot_list$Collaboration_span
}
Run the code above in your browser using DataLab