Introduction:

Considering that, in subjects such as statistics, affective aspects can determine the achievement of learning objectives, this study set out to establish the content and construct validity of the Scale of Attitudes towards Statistics.

Methods:

Cross-sectional observational study of scale validation type. 250 undergraduate students from health-related areas participated. Content validity was analyzed by expert judgment, construct validity with exploratory factor analysis and confirmatory.

Results:

In the content validation, all the items obtained scored higher than 3 (scale of 1-4), with optimal inter-judge agreement (> 0,5). Cronbach's alpha of 0,865. With the exploratory factor analysis 6/25 items were eliminated. The results of the confirmatory factor analysis reported an adequate Comparative Fit Index and acceptable Tucker-Lewis Index. The Root Mean Square Residual Error of Approximation was also acceptable (between 0,05 and 0,08), the Standardized Root Mean Square Residual indicated a good fit (< 0,08). The items were grouped into 3 factors: positive assessment of the statistics, liking and skills in statistics, negative assessment of the statistics.

Conclusion:

The Scale of Attitudes towards Statistics, validated in the Colombian population, is a support tool in the teaching-learning process of statistics; it can be used at the beginning, during follow-up or at the end of subjects that involve statistics.