Test-retest reliability of Attention Network Test measures in schizophrenia

The Attention Network Test (ANT) is a well established behavioral measure in neuropsychological research to assess three different facets of selective attention, i.e., alerting, orienting, and conflict processing. Although the ANT has been applied in healthy individuals and various clinical populations, data on retest reliability are scarce in healthy samples and lacking for clinical populations. The objective of the present study was a longitudinal assessment of relevant ANT network measures in healthy controls and schizophrenic patients.

Forty-five schizophrenic patients and 55 healthy controls were tested with ANT in a test-retest design with an average interval of 7.4 months between test sessions. Test-retest reliability was analyzed with Pearson and Intra-class correlations.

Healthy controls revealed moderate to high test-retest correlations for mean reaction time, mean accuracy, conflict effect, and conflict error rates. In schizophrenic patients, moderate test-retest correlations for mean reaction time, orienting effect, and conflict effect were found. The analysis of error rates in schizophrenic patients revealed very low test-retest correlations.

The current study provides converging statistical evidence that the conflict effect and mean reaction time of ANT yield acceptable test-retest reliabilities in healthy controls and, investigated longitudinally for the first time, also in schizophrenia. Obtained differences of alerting and orienting effects in schizophrenia case-control studies should be considered more carefully. The analysis of error rates revealed heterogeneous results and therefore is not recommended for case control studies in schizophrenia.