TY - JOUR
T1 - Sample length affects the reliability of language sample measures in 3-year-olds
T2 - Evidence from parent-elicited conversational samples
AU - Guo, Ling Yu
AU - Eisenberg, Sarita
N1 - Publisher Copyright:
© 2015 American Speech-Language-Hearing Association
PY - 2015
Y1 - 2015
N2 - Purpose: The goal of this study was to investigate the extent to which sample length affected the reliability of totanumber of words (TNW), number of different words (NDW), and mean length of C-units in morphemes (MLCUm) in parent-elicited conversational samples for 3-year-olds. Method: Participants were sixty 3-year-olds. A 22-min language sample was collected from each child during free play with the parent in the laboratory. Samples of 1, 3, 7, and 10 min were extracted from the 22-min samples. TNW, NDW, and MLCUm were computed from each shorter sampland the 22-min sample. TNW and NDW were adjusted by number of minutes for comparisons. The differences and correlations between each shorter sample cut and the 22-min sample on MLCUm and adjusted TNW and NDW were computed. Results: The shorter samples and the 22-min samples significantly differed in adjusted TNW and NDW, but not in MLCUm. TNW reached an acceptable reliability level (i.e., r = .90) in 7-min samples. NDW and MLCUm approached the acceptable reliability level (rs = .88) in 7-min samples and reached it in 10-min samples. Conclusion: For conversational language samples with similar collection procedures, samples of 7 to 10 min are desirable for calculating TNW, NDW, and MLCUm in 3-year-olds.
AB - Purpose: The goal of this study was to investigate the extent to which sample length affected the reliability of totanumber of words (TNW), number of different words (NDW), and mean length of C-units in morphemes (MLCUm) in parent-elicited conversational samples for 3-year-olds. Method: Participants were sixty 3-year-olds. A 22-min language sample was collected from each child during free play with the parent in the laboratory. Samples of 1, 3, 7, and 10 min were extracted from the 22-min samples. TNW, NDW, and MLCUm were computed from each shorter sampland the 22-min sample. TNW and NDW were adjusted by number of minutes for comparisons. The differences and correlations between each shorter sample cut and the 22-min sample on MLCUm and adjusted TNW and NDW were computed. Results: The shorter samples and the 22-min samples significantly differed in adjusted TNW and NDW, but not in MLCUm. TNW reached an acceptable reliability level (i.e., r = .90) in 7-min samples. NDW and MLCUm approached the acceptable reliability level (rs = .88) in 7-min samples and reached it in 10-min samples. Conclusion: For conversational language samples with similar collection procedures, samples of 7 to 10 min are desirable for calculating TNW, NDW, and MLCUm in 3-year-olds.
UR - http://www.scopus.com/inward/record.url?scp=84926289343&partnerID=8YFLogxK
U2 - 10.1044/2015_LSHSS-14-0052
DO - 10.1044/2015_LSHSS-14-0052
M3 - Article
C2 - 25615272
AN - SCOPUS:84926289343
SN - 0161-1461
VL - 46
SP - 141
EP - 153
JO - Language, speech, and hearing services in schools
JF - Language, speech, and hearing services in schools
IS - 2
ER -