Estimation of Premorbid Intelligence Quotient of the Intelligence Scale for Chinese Adult

唐细容,姚树桥
DOI: https://doi.org/10.16128/j.cnki.1005-3611.2012.06.010
2012-01-01
Abstract:Objective: To explore estimating IQ of the Intelligence Scale for Chinese Adult(ISCA).Methods: Standardized sample of the ISCA(n=2035) was roughly equally divided,at random,into two groups,one for development of the estimating formulae and the other for validation.A series of hierarchical linear regression analyses were carried out to develop formulae for estimating intelligence quotient.In the procedure,demographic variables,including age,education,gender,vocation,ethnicity and region of the country,as well as performance of 'hold' subtests were predictors.To validate those formulae,comparison between the estimated IQ's and actual IQ's were executed.Results: There was no significant difference between the demographic characteristics and test scores of the two groups.Ten estimating formulae were developed,of which adjusted R2 was 38.3%-86.6%,estimating standard error was 11.84-5.47.There was no significant difference between the mean of estimated IQ's and actual IQ's,while ranges of estimated IQ's were more limited than those of actual IQ's in different measures.Estimated and actual values showed high correlation of 0.62-0.93 and shared same IQ grades at ratio of 49.9%-67.8%.Estimation of combining performance of subtest and demographic variables showed more satisfactory validitied than that of demographic variables simply.However,the more subtests employed,the less each subtest contribute.Conclusion: Based on the data of the ISCA standardization sample,ten ISCA IQ estimating formulae were developed,they all show satisfied validity of estimation in intact adults.
What problem does this paper attempt to address?