com.itemanalysis.psychometrics.reliability

Interface ScoreReliability

All Known Implementing Classes:

AbstractScoreReliability, CoefficientAlpha, FeldtBrennan, FeldtGilmer, GuttmanLambda, KR21, RajuBeta

public interface ScoreReliability

Author:

J. Patrick Meyer

Method Summary

Modifier and Type	Method and Description
double[]	confidenceInterval() Confidence interval for the reliability estimate computed using the F-distribution.
java.lang.String	confidenceIntervalToString(double[] confidenceInterval) Creates a String representation of the confidence interval.
ScoreReliabilityType	getType() Type of reliability estimate
double[]	itemDeletedReliability() An array of reliability estimates without the item indexed by the position in the array.
java.lang.String	printItemDeletedSummary(java.util.ArrayList<VariableAttributes> var) A String representation of all item deleted reliability estimates.
double	totalVariance() Total observed score variance.
double	value() Estimates reliability.

Method Detail

itemDeletedReliability

double[] itemDeletedReliability()

An array of reliability estimates without the item indexed by the position in the array. For example, rel[0] = is the reliability estimate without the item at position 0. Similarly, rel[1] = is the reliability estimate without the item at position 1.

Returns:

value

double value()

Estimates reliability.

Returns:

estimate of reliability

printItemDeletedSummary

java.lang.String printItemDeletedSummary(java.util.ArrayList<VariableAttributes> var)

A String representation of all item deleted reliability estimates.

Parameters:

var - VariableAttributes that provide the variable names.

Returns:

getType

ScoreReliabilityType getType()

Type of reliability estimate

Returns:

type of reliability estimate.

confidenceInterval

double[] confidenceInterval()

Confidence interval for the reliability estimate computed using the F-distribution. This computation is only correct for Coefficient Alpha because the sampling distribution for other reliability estimates is unknown. As such, this method returns an approximation, at best, of the reliability estimate for all reliability estimates other than coefficient alpha. Note that the confidence interval is computed using the largest sample size in the covariance matrix. This value would be the sample size for pairwise deletion.

Returns:

a confidence interval for the reliability estimate.

confidenceIntervalToString

java.lang.String confidenceIntervalToString(double[] confidenceInterval)

Creates a String representation of the confidence interval. It is used for displaying results.

Parameters:

confidenceInterval - an array with the lower [0] and upper [1] bounds of the confidence interval

Returns:

a String representation of the confidence interval.

totalVariance

double totalVariance()

Total observed score variance. It is the sum of all values in the covariance matrix.

Returns: