Summarise Kappa Coefficients

Using a list of PDD algorithms stratified by IADL deficit operationalisation and a table of pairwise concordance statistics, this function extracts the raw Cohen’s kappa estimates and computes their summary statistics (mean and standard deviation) across groups of algorithm pairs defined by the IADL operationalisation scheme.

Usage

summarise_kappa(algorithms, concordance)

Arguments

algorithms

A list containing algorithm mapping generated by list_algorithms.

concordance

A tibble containing concordance indexes for pairs of algorithms generated by decribe_concordance. The table is also available as data concords.

Value

A list with the following components:

faq_tot

A numeric vector containing all Cohen's kappas of algorithm pairs with IADL deficit defined as FAQ total > 7.

faq_9

A numeric vector containing all Cohen's kappas of algorithm pairs with IADL deficit defined as FAQ item 9 > 1.

cross

A numeric vector containing all Cohen's kappas of algorithm pairs with different IADL deficit definitions.

iadl

A numeric vector containing all Cohen's kappas of algorithm pairs which differ only in IADL deficit definition.

sums

A named character vector containing mean and standard deviation of each vector of Kappas contained in this list.

See also

• list_algorithms() prepares algorithms.

• describe_concordance() prepares concordance.

• concords containg data used as concordance.

Examples

if (FALSE) { # \dontrun{
p <- data_paths("data-raw")
data <- prepare_data(p)
pdd  <- diagnose_pdd_sample(data)
vars <- here::here("data-raw", "VariablesOfInterest.csv")

data(concords)
rates <- summarise_rates(pdd, vars)
algos <- list_algorithms(rates$table)

# Two equally valid ways to run the function:
kappa1 <- summarise_kappa(algos, rates$table)
kappa2 <- summarise_kappa(algos, concords)
} # }