solve

computer-intensive methods / data transformation / distribution-free tests / nonparametric tests / parametric tests / randomization tests / rank-order tests / rank-order transformation / square-root transformation The previous analyses using the t-tests, analysis of variance, and the significance tests for correlation and prediction studies all assumed that the variables being compared were normally distributed in the population and had equal variances. Therefore, these tests are called (1) . Since the outlying scores observed by the psychologists are high scores, the distributions of scores are skewed to the right. Due to the nature of their data, the psychologists decide to adjust the scores so that the distributions appear more like normal distributions by making a (2) . With a dataset that is skewed to the right, an adjustment that often works well is a (3) . If the psychologists arrange the scores in order from lowest to highest, they will be making a (4) . One advantage of this adjustment is that the psychologists could use the familiar statistical tests used earlier. Alternatively, the adjustment would permit the use of different tests called (5) Since the transformation results in a rectangular distribution, the psychologists do not need to estimate values like the population variance. Therefore, these tests are called (6) . Since the shape of the score distribution does affect the results of the analysis, these tests are also called (7) . An additional alternative would be for the psychologists to use newer methods called (8) . These procedures involve combining sources into all possible groupings based on chance, so they are called (9)