Psychological Test Construction: Teacher’s Caring

Psychological Test Construction: Teacher’s Caring

Muhammad Aji Winata^a, Nurul Arbiyah^b*, and Gloryka Ednadita^c*

^aFaculty of Psychology, Universitas Indonesia, Depok, Indonesia; ^bPsychology Research Method Departement, Faculty of Psychology, Depok, Indonesia; ^cPsychology Research Method Departement, Faculty of Psychology, Depok, Indonesia

*Corresponding Author:

Nurul Arbiyah

Psychology Research Method Department

Faculty of Psychology, Universitas Indonesia

Jl. Lkr. Kampus Raya, Depok, Jawa Barat

Indonesia, 16424

Tel.: +62 217270004

Email address: nurul.arbiyah31@ui.ac.id

Gloryka Ednadita

Psychology Research Method Department

Faculty of Psychology, Universitas Indonesia

Jl. Lkr. Kampus Raya, Depok, Jawa Barat

Indonesia, 16424

Tel.: +62 217270004

Email address: ednadita.gloryka@ui.ac.id

Psychological Test Construction: Teacher’s Caring

Caring is one of the essential characteristics for an effective teacher. Without a caring teacher, the learning process will not happen optimally and can affect the student’s well being. However, research done related to this construct is still very few in Indonesia. One of the reason because there are no measuring tools related to caring in Indonesia, so it affects the amount of research about caring. Based on that, this study aims to construct a Teacher Caring Scale (TCS) for secondary school teachers that is reliable, valid, has good items, and norm to interpret the scores.

The participants of the study were secondary school teachers in Jakarta, Bogor, Depok, Tangerang, and Bekasi (N = 291). For data collection, we use non-probability sampling, specifically convenience sampling for reasons of ease, accessibility, and closeness of the participants with the researchers. There are two instruments used in this study; Self-report Teacher Caring Scale that we developed based on McCroskey and Teven’s (1997) theory and Big Five Inventory: Agreeableness for purpose of construct validity testing. TCS can be categorized as typical performance test and it uses a Likert scale to measure response from the participants. There are 42 items tested in the field data collection and there will be 27 items for the final TCS measuring instrument. Scores obtained from the participants will illustrate the level of caring owned by the teacher. The higher the score that teacher get, it indicates the higher caring level that is owned by the teacher. The present study is done to examine the psychometric qualities of Teacher Caring Scale. Reliability testing is examined by using Cronbach alpha, while validity testing is examined by using several construct validity tests, such as convergent validity and confirmatory factor analyis. Further analyses of the items are corrected item-total correlation (Crit) and factor loading.

The results of the psychometric test with Cronbach alpha indicates that TCS has a good internal consistency (α = 0.916). Construct validity shows significant correlation with agreeableness (r = 0.371, p <0,01). This result indicates that TCS is valid for measuring teacher's caring. Based on the internal consistency validity test, the components in the TCS are significantly correlated with the total TCS score that indicates good internal consistency. The data also has a good fit with the thereotical model, p-value = 0.062, CFI = 0.97, GFI = 0.88, and RMSEA = 0.048. The results of item analysis using item discrimination (Crit) indicates that items on the TCS can differentiate teachers who have high caring with teachers who have low caring. The result of item analysis using a factor loading based on CFA indicates that the items on the TCS measuring instrument have contributed well to the measured latent variable which is teacher’s caring. However, the results of the proportion of endorsement analysis of some items indicate extreme scores that tend to provoke responses approved or disapproved by the participants. Nevertheless, response variation on some items is still well distributed. Finally, the TCS test's score is interpreted using a normalized standard score with (M = 10, SD = 3).

These results indicate that the TCS measuring instrument is a useful test because it has good reliability and validity. The final TCS measurement consists of 27 items that can adequately discriminate teacher’s caring and has factor loading that contributes to TCS measurement. As caring is one of the qualities that can improve the teacher’s competence, this tool can be useful in an effort to improve the competence of teachers that can be utilized by each school, education authorities, and by the ministry of education. This measurement can be used to determine baseline of the teacher’s caring and as one of the references in the implementation of teacher development program.

Keywords:

pychometric testing, teacher’s caring, Teacher Caring Scale (TCS)

Word Count: 618