[Configuration frequency analysis: Kieser's test against Victor's type and syndrome].

Tests against types and anti-types as performed in Victor's Configural Frequency Analysis (CFA) have shown to be non-conservative. The conservative test by Kieser is introduced. It is applied to data describing a sample that performs below average in language tests. For data including small expected cell frequencies the Harms-Ihm adjustment is recommended. It leads to valid chi-squared testing even if expectancies are small. The problem of identifying more than two types or syndromes in exploratory CFA is discussed.