Four possible cases in trend analyses are discussed: unequal n-unequal intervals, unequal n-equal intervals, equal n-unequal intervals, equal n-equal intervals. The presence of unequal n (with equal or unequal intervals) usually produces nonorthogonality of components, necessitating corrective procedures. The suggested procedure is the use of n in equations to derive the orthogonal polynomial coefficients. Unequal intervals (with equal or unequal n) requires that the orthogonal polynomial coefficients be derived because tabled coefficients are not appropriate.
Get full access to this article
View all access options for this article.
References
1.
Boneau, G. A.The effect of violations of assumptions underlying the t test. Psychological Bulletin , 1960, 57, 49-65.
2.
Gaito, J.Unequal intervals and unequal n in trend analyses. Psychological Bulletin, 1965, 63, 125-127.
3.
Gaito, J.An index of estimation to ascertain the effect of unequal n on anova F tests. American Psychologist, 1972, 27, 1081-1082.
4.
Gaito, J.Introduction to analysis of variance procedures. New York: MSS Information Corp., 1973.
5.
Hays, W. L.Statistics for psychologists. New York: Holt, Rinehart, and Winston , 1963.
6.
Kirk, R. E.Experimental design: procedures for the behavioral sciences. Belmont, Calif.: Brooks/Cole, 1968.
7.
Lewis, D.Quantitative methods in psychology . New York: McGraw-Hill, 1960.
8.
Winer, B. J.Statistical principles in experimental design. New York: McGraw-Hill, 1971.
