MGMT969 Chapter 3 part 2
Home > Flashcards > Print Preview
The flashcards below were created by user
on FreezingBlue Flashcards. What would you like to do?
Measurement and individual differences. This example highlights the important role of measurement in selection research. Numbers assigned to predictors and criteria enable us to make the necessary fine distinctions among individuals. Subsequent analyses of these numbers help us meet one of our goals: devveloping a system for predicting job performance successfully. Without measurement, we would probably be left with our intuition and personal best guesses. Perhaps, for a very small number of us, these judgments will work. For the most - deciding by the seat of our pants will not work.
Measurement is to identify those who should be hired for the job. Predictors and criteria help identify those who should be hired. e.g. making baskets - very few people will make really high quantity or low quantity (most will make somewhere in the middle - bell curve). If we can assume quantity of production is a suitable criterion , our objective is to obtain a predictor that will detect the individual difference or variance in productivity. I If our predictor is useful, individuals' scores on the predictor will be associated with their productivity scores.
scales of measurement. The use of predictors and criteria in selection research requires that these variables be measured - rigorously. This is because our ability to distinguish one person from another is determined by the precision with which we measure the variables, rigorus measurement is a prerequisite for performing any statistical analysis in a selection study.
precision means the number of distinct scores or gradations permitted by the predictor an criterion used. The level of precision will dictate what statistical analyses can be done with the numbers obtained fromm measurement.- greater precision enables the use of more sophisticated statistical analyses.
scale of measurement
a means by which individuals can be distinguished from one another on a variable of interest, whether that variable is a predictor or criterion - variety of scales = specific predictor or criterion variables chosen can differ rather dramatically in their precision.
example of developing a selection program for bank management trainees.
- One criterion we want to predict is trainability - the trainee success in a management training program. Trainability can be measured in several ways. On one hand we could simply classify individuals according to who did and did not graduate from the training program (graduation, for example, may be based on trainees' ability to pass a test on banking principles and regulations.) Our criterion would be a dichotomous category, that is, unsuccessful (fail) and Successful (pass). Thus our predictor variable would be used simply to differentiate between those applicants who could and those who could not successfully complete their training.
- another way to evaluate trainability is by the degree of trainee succes as measured by training performance test scores. The test might be used to assess how much trainees know at the end of their training program. In this case the criterion is not a categorical measurement(ie. success or failure) but a way to more precisely describe the degree of success experienced in training.
classification of success criterion vs degree of success criterion
classification of success is not measured as precisely as the degree of success criterion (on the graph). Greater individual differences in trainability can be mapped for the degree of success criterion. The variable trainability is the same in both but the two differ with respect to the level of measurement involved. It is the manner in which the criterion is measured and not the criterion itself that determines level of measurement. One measure of trainability can dram more precise conclusions than the other.
four types of scales or levels of easurement exist
- A - nominal
- B - ordinal
- C - interval
- D - ratio
- The degree of precision of measuring differences among people increases as we move from nominal to ratio scales.
- Increased precision proides use with more detailed information about people with regard to the variables being studied.
- Whether those variables are employee attributes such as mental ability and personality or criteria such as trainability and performance. More powerful statistical analyses can then be performed with the data.
Nominal scale is composed of two of more mutually exclusive categories. - generally information is statistical
- The numerical values serve merely as 'label' for the category and do not imply any ordering of the attributes. E.g. coding male applicatns "1" and Female applicants "2" does not signify females are twice whatever males are. These numbers although meaningful and useful, carry no numerical rating.
- We could use other numbers as well (such as 0 and 1) to identify females and males. Both scoring schemes have the same meaning. The numerical codes themselves do not indicate how males and females differ. The only information we have through our numerical codes is an applicant's sex. Because we cannot state how members assigned to nominal scale categories differ, this type of scale is the simplest form of measurement.
What would you like to do?
Home > Flashcards > Print Preview