Methodological problem of selection of the unit of analysis for the studying prenatal X-irradiation effects on open-field behavior in rats
スポンサーリンク
概要
- 論文の詳細を見る
Sprague-Dawley rats were given X-irradiation (150 R) on Day of gestation. Male offspring were administered an open-field test at about seven weeks of age. Magnitude of dependence among scores of littermates was estimated in terms of correlation coefficient, using data from the openfield test and brain weight. The <I>r</I>s in ambulation, rearing and penetration into the inner square (PIS) were somewhat large in the X-irradiated group. The <I>r</I>s for brain weight was very large. The values in the control group were relatively small and close to zero. The values for littermates postnatally constructed were also small in both groups fostered by X-irradiated mother or control mother, except for brain weight. The magnitudes of type I error rate distortions when individual scores are used instead of litter scores as the basic unit of statistical analysis were assessed. The type I error distortions in ambulation, rearing and PIS were somewhat large in the prenatal experimental group. The remaining open-field measures and all open-field measures for postnatal groups were relatively near to the nominal values. The type I error distortion in brain weight was very large. Therefore, one should use litter as a basic unit in the analysis of brain weight. As to open-field measures, it might be advisable to use litter whenever prudence is especially required. However, the type I error distortion by using subject as a unit is not serious.
- 公益社団法人 日本心理学会の論文