Developing a physics laboratory anxiety scale

The purpose of the research

This research was conducted to develop a scale which can be applied to determine the anxieties related to physics laboratory lessons of students from secondary schools and higher education institutions.

Study group

The study group consisted of 245 undergraduates taking physics laboratory lessons during the 2011-12 academic year at Necmettin Erbakan University, Ahmet Keleşoğlu Faculty of Education, in Turkey.

They were undergraduates in physics, chemistry and science and their ages ranged from 17 to 20. 67 (27.3%) undergraduates in the sample were male and 178 (72.7%) were female.

Development Process

1. Forming the design of the scale: in order to determine the items required for a physics laboratory anxiety scale, 20 undergraduates taking physics laboratory lessons were asked to write down any anxieties they had about the physics lesson. These were then discussed. The item pool, which was based on these answers, was examined by seven physics lecturers and the 42-item scale design was formed.

2. Content validity: content validity is the indicator which shows whether items are qualitatively and quantitatively adequate for measuring the behaviour that is to be measured (Büyüköztürk, 2007). Expert opinions were employed in order to determine the content validity. Besides the seven physics academics whose opinions were used during the scale design, an assessment and evaluation expert and a Turkish teaching academic were consulted. After implementation of the recommended changes, a 42-item scale design was completed.

3. Application: the 42-item scale, which is of five-point Likert type, was tested on 245 undergraduates in the sample group in the spring semester of the 2011-12 academic year.

4. Construct validity analysis: construct validity shows the degree to which a test correctly measures an abstract concept in terms of the behaviour that is to be measured (Büyüköztürk, 2007). Exploratory and confirmatory factor analyses were performed in order to examine the construct validity of the scale.

5. Reliability analysis: to test the reliability of the scale, the Cronbach's alpha reliability coefficient was calculated. The Cronbach's alpha reliability coefficient is an indicator of internal consistency between the test points of the scale. If Cronbach ? reliability is 0.70 or higher, the reliability of the instrument is adequate (Büyüköztürk, 2007).