PEDS109 Lecture Notes - Frequency Distribution, Abscissa And Ordinate, Multimodal Distribution
Document Summary
When we go with a grouped frequency distribution table, we lose information because we are no longer dealing with raw scores. The more rows we have, the less loss of info and vice versa. 500,000 and 1,000,000 are good numbers to use as interval widths. Choose the number of rows that are closer to 10 (8 will be closer than 16 to 10) If we are working with one million intervals, the bottom score in each class interval should be a multiple of the width. Extremely skewed distribution lots of people on the high end, few people on the low end. Have the numbers in ascending or descending orders. Rounding error (we wil always work to 3 decimal places in this course) Use these only if you are using ratio measure. There is no space beween the bars, that tells the statistician that you are working with interval or ratio measurement.