The Impact of Guessing on the Accuracy of Estimating Simple Linear Regression Equation Parameters and the Ability to Predict

Abstract

The current study aimed at identifying the impact of guessing on the accuracy of estimating simple linear regression equation parameters and the ability to predict. To achieve the objectives of the study, an achievement test was built in its final form of (40) multiple-choice items in a measurement and evaluation course. After verifying the psychometric properties of the test, it was applied to a sample of (134) male and female general diploma students registered at AL-Hussein Bin Talal University in the second semester of the year 2020/2021. The test was divided into two sections: the first section was tested based on the traditional correction method, while the second section was tested based on the correction formula for the impact of guessing. The results of the study concluded that there are statistically significant differences among the predicted values in the scores of the examinees when using the traditional method and the correction formula for the impact of guessing in favor of the traditional method, and that there are statistically significant differences among the means of residual squares for estimating the scores of the study sample attributed to the correction method and in favor of the traditional method. The results of the study also found that the value of the explained variance (R2) increases when using the correction equation for the guessing effect, and that the values of the parameters of the simple linear regression equation increase when using the correction equation for the impact of guessing, and that the values of the standard error in estimating the parameters of the simple linear regression equation decrease when using the correction equation for the impact of guessing.