Pregled bibliografske jedinice broj: 1255260
How to integrate the information on accuracy, reaction time and confidence rating in the psychometric analysis of test items?
How to integrate the information on accuracy, reaction time and confidence rating in the psychometric analysis of test items? // 21. Dani Ramira i Zorana Bujasa - knjiga sažetaka / Kuterovac Jagodić, G. ; Erceg Jugović, I. ; Huić, A. (ur.).
Zagreb: Odsjek za psihologiju Filozofskog fakulteta Sveučilišta u Zagrebu, 2013. str. 64-64 (predavanje, domaća recenzija, sažetak, znanstveni)
CROSBI ID: 1255260 Za ispravke kontaktirajte CROSBI podršku putem web obrasca
Naslov
How to integrate the information on accuracy, reaction time and confidence rating in the psychometric analysis of test items?
Autori
Ljubotina, Damir
Vrsta, podvrsta i kategorija rada
Sažeci sa skupova, sažetak, znanstveni
Izvornik
21. Dani Ramira i Zorana Bujasa - knjiga sažetaka
/ Kuterovac Jagodić, G. ; Erceg Jugović, I. ; Huić, A. - Zagreb : Odsjek za psihologiju Filozofskog fakulteta Sveučilišta u Zagrebu, 2013, 64-64
ISBN
978-953-175-463-7
Skup
21. Dani Ramira i Zorana Bujasa
Mjesto i datum
Zagreb, Hrvatska, 11.04.2013. - 13.04.2013
Vrsta sudjelovanja
Predavanje
Vrsta recenzije
Domaća recenzija
Ključne riječi
computer based testing, accuracy, reaction time, confidence rating
Sažetak
Computer based testing enabled conceptually new approaches in test administration as well as new information associated with the problem solving process and response choice. Relationship between reaction times and scores on mental ability tests was the main focus of research efforts in the area of cognitive psychology for a long time. Results have been widely reported in literature on the information processing theories of intelligence but there have been varying estimates of the strength of these relationships. More recent research is focused on general models attempting to include reaction times into some of the existing item response theory models or on modeling reaction times independently of other item performance parameters. In this study we attempted to assess different item performance measures and their potential correlates in order to propose an integrated item response model. During computer based administration of knowledge tests we recorded response accuracy as well as different variables related to time spent on the item: first response time, last response time (if the respondents changed their answers) as well as the total exposition of the item. Respondents also rated the degree of their confidence in their responses. As potential correlates of the dependent variables used for item performance we assessed choice reaction time and general cognitive processing speed, as well as scores on questionnaires on impulsivity, conscientiousness, perfectionism, temperament, test anxiety, self-esteem and meta-cognitive aspects related to the constructs measured by the test. Preliminary results point to significant possibilities for additional insight into the process of item response. The results indicate complex relationships between the measures used, where the relationship between speed and accuracy is moderated to a greater extent by item characteristics than with personality traits of the respondents.
Izvorni jezik
Engleski
Znanstvena područja
Psihologija