Test-retest reliability of the new cognitive preschool screening tests

Date of Award




Degree Name

Doctor of Philosophy (Ph.D.)


Developmental Psychology

First Committee Member

Marcia S. Scott, Committee Chair


The demand for more effective screening tests to identify mild learning impairments in young children is noted throughout the literature. Two new screening tests have been developed to target prekindergarten (PK) and kindergarten (KG) children with mild cognitive impairments such as learning disabilities (LD) or mild mental retardation (MMR). An evaluation of the test-retest reliability of these new preschool screening tests is a necessary requirement since test reliability will impact test validity. One hundred and sixty-eight children were administered either the prekindergarten (N = 82) or the kindergarten (N = 86) cognitive screening test within a two week interval. Three measures of reliability, Pearson's correlation, Cronbach's alpha, and Kappa coefficient, were used to examine the test-retest reliability of the prekindergarten and Kindergarten screening tests. In general, the results indicated high levels of reliability for the PK test with moderate levels for the KG test. Level of stability as measured by Pearson's correlation indicated a value of .85 for the PK sample and .75 for the KG group. Cronbach's alpha or level of internal consistency were moderate for both scales, although lower for the KG scale at time two. An assessment of the consistency of classification (below vs. above the cutoff score) was reflected in a Kappa coefficient of .8 for PK but only .4 for the KG group. In summary, the PK screening test is reliable but for the KG test, reliability is less than optimal.


Education, Early Childhood; Psychology, Developmental; Psychology, Psychometrics; Psychology, Cognitive

Link to Full Text