'Hartmann distance' (standardized Slater distances).

Calculate Hartmann distance

Usage

distanceHartmann(
  x,
  method = "paper",
  reps = 10000,
  prob = NULL,
  progress = TRUE,
  distributions = FALSE
)

Arguments

x: repgrid object.
method: The method used for distance calculation, on of "paper", "simulate", "new". "paper" uses the parameters as given in Hartmann (1992) for calculation. "simulate" (default) simulates a Slater distribution for the calculation. In a future version the time consuming simulation will be replaced by more accurate parameters for Hartmann distances than used in Hartmann (1992).
reps: Number of random grids to generate sample distribution for Slater distances (default is 10000). Note that a lot of samples may take a while to calculate.
prob: The probability of each rating value to occur. If NULL (default) the distribution is uniform. The number of values must match the length of the rating scale.
progress: Whether to show a progress bar during simulation (default is TRUE) (for method="simulate"). May be useful when the distribution is estimated on the basis of many quasis.
distributions: Whether to additionally return the values of the simulated distributions (Slater etc.) The default is FALSE as it will quickly boost the object size.

Value

A matrix containing Hartmann distances. In the attributes several additional parameters can be found:

arguments: A list of several parameters including mean and sd of Slater distribution.
quantiles: Quantiles for Slater and Hartmann distance distribution.
distributions: List with values of the simulated distributions.

Details

Hartmann (1992) showed in a simulation study that Slater distances (see distanceSlater()) based on random grids, for which Slater coined the expression quasis, have a skewed distribution, a mean and a standard deviation depending on the number of constructs elicited. He suggested a linear transformation (z-transformation) which takes into account the estimated (or expected) mean and the standard deviation of the derived distribution to standardize Slater distance scores across different grid sizes. 'Hartmann distances' represent a more accurate version of 'Slater distances'. Note that Hartmann distances are multiplied by -1. Hence, negative Hartmann values represent dissimilarity, i.e. a big Slater distance.

There are two ways to use this function. Hartmann distances can either be calculated based on the reference values (i.e. means and standard deviations of Slater distance distributions) as given by Hartmann in his paper. The second option is to conduct an instant simulation for the supplied grid size for each calculation. The second option will be more accurate when a big number of quasis is used in the simulation.

It is also possible to return the quantiles of the sample distribution and only the element distances considered 'significant' according to the quantiles defined.

Calculation

The 'Hartmann distance' is calculated as follows (Hartmann 1992, p. 49).
$$D = -1 (\frac{D_{slater} - M_c}{sd_c})$$ Where $D_{slater}$ denotes the Slater distances of the grid, $M_c$ the sample distribution's mean value and $sd_c$ the sample distribution's standard deviation.

References

Hartmann, A. (1992). Element comparisons in repertory grid technique: Results and consequences of a Monte Carlo study. International Journal of Personal Construct Psychology, 5(1), 41-56.

Examples

if (FALSE) { # \dontrun{

### basics  ###

distanceHartmann(bell2010)
distanceHartmann(bell2010, method = "simulate")
h <- distanceHartmann(bell2010, method = "simulate")
h

# printing options
print(h)
print(h, digits = 6)
# 'significant' distances only
print(h, p = c(.05, .95))

# access cells of distance matrix
h[1, 2]

### advanced  ###

# histogram of Slater distances and indifference region
h <- distanceHartmann(bell2010, distributions = TRUE)
l <- attr(h, "distributions")
hist(l$slater, breaks = 100)
hist(l$hartmann, breaks = 100)
} # }