Programs NPCA1 and NPCA2 are developed to analyze experimental data consisting of repeated measurements on the same individuals at consecutive time points within a certain observation period.

Program NPCA1 is designed to analyze “temporary increase” type curves, i.e., when values increase and then return to the baseline. (By changing the sign of the values, it can be applied even when a temporary decrease is expected.) For each individual curve, the procedure identifies the “top period”, where the highest values occur, with some “transient periods” around it and tests whether the increase is significant.

Program NPCA2 is designed to analyze more general responses (without returning to the baseline). The analysis focuses on the time and value of the minimum and maximum, so it is most appropriate if primarily these measures are of interest. For each individual curve, the procedure determines those two periods in which the measurement under study has maximal and minimal average value (referred to as the “max period” and the “min period”) and tests whether the difference between the average values of these two periods is significant.

Program NPCA-ADD combines the individual p-values into an overall (sample) p-value by Edgington’s additive method, while program NPCA-TIM tests whether the times of maximum (or minimum) are distributed uniformly in the observation period (against accumulation in some parts of it) by applying a randomization version of the chi-square test.

For further details of the procedure, see the following papers:

Reiczigel J, Analysis of data with repeated measurements, Biometrics, 55, 1059-1063 (1999).

Reiczigel J, Bajcsy AC, Szenci O, A simple nonparametric analysis for longitudinal data, Conference poster, Joint conference of ISCB and GMDS, Heidelberg, 1999.

For an application of NPCA2, see

Bajcsy AC, Reiczigel J, Szenci O, Circadian changes in blood ionized calcium, sodium, potassium, and chloride concentration and pH in cattle, American Journal of Veterinary Research, 60, 945-948 (1999).

For technical details on the usage of the programs, see the readme file.

Here you can download a zip archive (376 k) containing all programs, the readme file, and some example data and outputs.