Premium
Variability of classification by the matching familiar figures test as a function of grade level medians
Author(s) -
Ames Steven G.
Publication year - 1977
Publication title -
psychology in the schools
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.738
H-Index - 75
eISSN - 1520-6807
pISSN - 0033-3085
DOI - 10.1002/1520-6807(197707)14:3<382::aid-pits2310140322>3.0.co;2-4
Subject(s) - median , matching (statistics) , psychology , cutoff , statistics , test (biology) , mathematics , paleontology , physics , geometry , quantum mechanics , biology
This study examined the variability of classification by the Matching Familiar Figures Test (MFF) as a function of grade level medians. The MFF was administered to 52 kindergarten and 60 first‐grade children, and classification was completed using two sets of medians. The first procedure involved classification based on medians for response time and errors computed from the combined MFF scores of the kindergarten and first‐grade children. The second procedure involved clmsification of the same children based on separate medians for each grade. A x 2 analysis indicated that MFF classification was highly dependent upon which sample medians were used; 18%(N = 20) of the sample failed to maintain the same classification across both procedures. Analysis also revealed that for the majority (65%) of the children change in classification was due to the change in the median error cutoff. The implications of these findings for future MFF research involving young children and the need for standardized norms for the MFF are discussed.