| The effect of positive and negative wording on the scale is called wording effect.The researchers developed the mixed form personality Questionnaires containing positive and reverse expressions in order to prevent the tendency of respondents to agree with the content of the questions and not serious behavior,so as to reduce the response bias of the subjects.With the popularity of computer-based testing(CBT),response time(RT)information has been collected.A large number of studies have shown that(Thissen 1983;van der Linden and Guo 2008;Ranger,2013),reaction time is a very important source of potential information,which provides information about individual potential traits and item characteristics,and helps to improve the accuracy of the estimation results of subjects’ traits.The wording effect implies the differences in cognitive processing strategies caused by different wording directions.Positively worded items had higher cognitive fluid than negative wording items(Bruckmuller & Hoorens,2015).Therefore,the cognitive processing of negative wording items is more complex and requires more reaction time(Clark,1976).In cognitive psychology,the time of information processing indicates the cognitive process of information processing(Schnipke &Scrams,2002).It can be seen that the reaction time information can be used to explore the influence of wording effect on the cognitive process of the participants.The fusion of wording effect and response times information can not only control the influence of item expression effect,but also use the dual information of the relationship between wording effect and response times,so as to improve the estimation accuracy of the subject’s trait parameters.At present,neither the item response model nor the reaction time model has a psychometric model that combines the information of wording effect and reaction times.Therefore,the purpose of this study is to develop an IRT model that combines the information of item presentation effect and reaction time to cope with the mixed-format personality scale(recorded as Bi-GPCM-RT).The model contains two sub-models,one is used to describe the wording effect information,and the other is used to describe the reaction times and wording effect relationship.The Markov chain Monte Carlo(MCMC)algorithm is used to estimate the model parameters.We carried out two simulation studies to verify the rationality of the model,and showed the application value of the model through the analysis of real data.The results are as follows:(1)The Monte Carlo simulation study shows that:Compared with the traditional GPCM and BIMFS models,the Bi-GPCM-RT model with wording effect and response times information developed in this study has higher parameter estimation accuracy of subjects regardless of the experimental conditions,which indicates that the combination of wording effect and response times information can effectively improve the estimation accuracy of subject trait parameters.(2)The empirical data study shows that the new model has a better fit to the real mixed-format personality test data than the traditional IRT model,which shows the rationality and advantages of the new model in the mixed-format personality test.The item response model combined with wording effect and reaction times information constructed in this study promoted the scientific application of mixed-format personality inventory in the field of item response theory,and was conducive to improving the estimation accuracy of subjects’ personality traits.In addition,this paper also discusses several potential application directions of the proposed model and finally introduces the research prospects for further improvement and perfection in the future. |
PDF Full Text Request