(Post last updated July 21, 2021)
Review panel summary
The Chemistry Laboratory Anxiety Instrument (CLAI) is a 20-item Likert scale (5-point strongly disagree to strongly agree scale) instrument that assesses students’ anxiety in the chemistry laboratory. During development, the CLAI was evaluated with 361 first and second year chemistry laboratory students from four US research institutions [1]. The final 20-item inventory was evaluated with 598 students from four institutions from across the United States [1]. Limited evidence of validity and reliability have been established for the CLAI. The primary means of validity that the authors present is the internal structure validity via the use of exploratory factor analysis suggesting that the CLAI can distinguish between 6 different aspects of laboratory anxiety in chemistry [2]. Additionally, the authors correlated the scale scores for each of the 6 factors with one another to provide evidence of the multidimensional nature of laboratory anxiety [1]. Additional correlations were made between the anxiety scales and students’ expected lecture and laboratory grades and most of the scales were found to negatively correlate with both expected grades providing validity evidence for the CLAI scores relation to other variables [2]. Finally, coefficient alpha was used to estimate the single administration reliability of the item groupings [1]. Using the data from all evidence of validity and reliability, the authors decided that the CLAI really only truly assesses 5 of the 6 factors and those are the ones included in the final 20-item instrument: 1) working with chemicals, 2) using equipment and procedures, 3) collecting data, 4) working with other students, and 5) having adequate time.
Recommendations for use
The CLAI was developed to measure undergraduate students’ anxiety in the chemistry laboratory in 5 areas: 1) working with chemicals, 2) using equipment and procedures, 3) collecting data, 4) working with other students, and 5) having adequate time. There was some evidence to suggest the existence of these 5 groupings, however the data presented suggests that some groupings may be questionable. Additionally, no evidence of the internal structure was provided for the final 20-item inventory, only the 30-item version that the authors began with. Due to the unclear nature of the groupings, it is recommended that CLAI instrument data be further investigated for evidence to support the internal structure. Additionally, the instrument was designed and administered as a 5-point Likert scale and as such should be used in this way.
Details from panel review
The CLAI developers provided limited evidence for validity and reliability of data collected with the instrument. It was unclear to the panel if the authors based the categories of anxiety on any theoretical or literature basis or just personal experience. As such, the test content validity of the instrument seems to be in question. The authors did conduct an exploratory factor analysis to provide evidence of internal structure validity following appropriate methods for the technique and found 6 factors; however some items loaded onto multiple factors and the loadings ranged from an absolute value of 0.50 to 0.80 [1]. The authors description of how they grouped items made many of the panel feel as if they were making the data fit the a priori categories for which the items were developed. To further support the multidimensional nature of the CLAI, the authors computed correlations between factors and found low correlations to establish distinct categories [1]. The authors used their determined structure to calculate coefficient alpha values to support their single administration reliability. The values ranged from 0.73-0.88 for the 6 factors which are viewed to be appropriate [1]. However, it is not clear that there is enough evidence that the 6 factors are measuring the constructs the authors propose. After presentation of all the evidence for 6 factors, the authors decided to remove the factor assessing “relief finishing the lab” as the authors indicate, the intended purpose of the CLAI is to measure different aspects of anxiety about chemistry laboratory learning, which resulted in a final 5-factor inventory [1]. The panel would like to note that there has been no evidence to support a 5-factor model for the instrument.
The CLAI scale scores do seem to somewhat correlate with students’ expected grades in lecture and lab, which provides some validity evidence for relation to other variables. However, it is important to note that the effect sizes that were reported for the correlations were small to null [1], so the panel did question to what degree this relation exists. Given the relation to affective constructs like anxiety and the large literature base in this area, the panel believes that the CLAI would be strengthened by further support in regard to response process validity and in understanding if the items developed do in fact measure the proposed aspects of anxiety,providing evidence of test content validity.
References
[1] Bowen, C. W. (1999). Development and Score Validation of a Chemistry Laboratory Anxiety Instrument (CLAI) for College Chemistry Students. Educational and Psychological Measurement, 59(1), 171–185. https://doi.org/10.1177/0013164499591012