measurement is essential to the successful implementation and administration of an HR selection program.
generally few applicants will be extremely bright and only a few extremely dull. What is needed to make finer distinctions among the rest with regard to the various characteristics that are of interest to us (intelligence, conscientiousness etc).
We need to use measurement to make discriminationsand to study in detail the relationship between applicant characteristics and employee performance on the job.
what is measurement?
hr perspective - measurement involves the systematic application of rules for assigning numerical values to objects to represent the quantities of a person's attributes or traits.
rules suggest that the basis for assigning numbers is clearly specified and consistently applied.
For any measures we might use in selection, eg. test - it is important that diifferent users who employ the measure obtain similar results. That is that the only differences in the test should relate to their ability to take the test and not differ according to the people assigning the score or the test-taking conditions.
Rules for assigning numbers to our selection measure help to standardize the application of these measures.
attributes are assessed and measured either through observation (e.g. physical) or non directly observable attributes known as psychological attributes.
Psychological attributes are not directly observable and include attributes such as conscientiousness, intelligence, job knowledge, and mathematical and verbal abilities (that must be inferred from an observable score - consisting of numbers or units of measurement). Obtaining a good measure of these elusive but critical psychological attributes is not always easy. - however psyhological traits will typically be identified as important indicators of how well an applicant can perform a job.
when we use tests or interviews to obtain measurements of psychological attributes, we must draw inferences from these measures. Because inferences are involved, we are on much shakier ground than when we can directly observe an attribute.
Someone conducting a selection interview might believe that the extent of an applican't eye contact with the interviewer reflects the applicants interest in the company or that the firmness of a handshake reflects the interviewee's self confidence- are they warrented? this question must be answered if we are to use numbers in a meaningful way to evaluate important psychological attributes when making selection decitions.
Third point in measurement definition
numbers or units represent attributes. Numbers play a useful role in summarizing and communicating the degree, amount or magnitude of an assessed attribute. Thus, if applicant A scores 40/50 on an achievement test and applicant B scores a 30 we conclude that applicant A has more 'achievement' than applicant B. Although these numbers signify degrees of difference in achievement, one must remember that achievement
itself is not measured difectly. it is instead inferred from the test score. Nevertheless, this score is used as an indicator of the construct. In this sense, numbers can provide a convenient means for characterizing and differentiating among job applicants. For this reason, numbers plan an important role in slection.
Fundamental challeng in personnel selection is choosing from a large group of job appicants the smaller group that is to be employed. The goal of personnel selection is to identify the individuals who should be hired. These individuals are chosen beccuase the selection measures predict that they will best perform the job in question.
When the term predict is used, it is assumed that a selection expert has the information on which to base these predictions.
where does someone get the prediction information?
Basically the information is derived from an analysis of the job.
is designed to identify the atributes incumbents must possess to perform the job successfully. Once we know the attributes necessary for success on the job, we can measure applicants with respect to these attributes and make our predictions accordingly.
Measurement plays a vital role in helping the selectionmanager make accurate predictions.
Predicting who should be hired generally requres the identification of two types of variables. Criterion-criteria and predictor variables.
usually serves as a definitiion of what is meant by employee success on the job. Criteria are defined by thoroughly studying the jobs fo which a selection system is being developed. A wide array of variables might serve as criteria. Some criteria deal with issues such as absenteesism, turnover, and other organisational citizenship behaviours. Other criteria represent work-related outcomes including error rates, number of goods produced, dollar slaes amount of scrap produced in manufacturing tasks, and speed of performance. This is just a small sampling of criteria that depict what some workers do. The most frequently used performance criteria are supervisory ratings of job performance. Training success and work samples are measured and used when they are meaningfully related to job performance.
Numberous types of criteria can be predicted -
Criteria should not be defined or chosen in a cavalier, usystematic or haphazard manner. They MUST be important to the job and MUST be appropriately measured. Becuase criteria are the basis for characterising employees success, the utility of a selection system will depend to a significant degree on their relevance, definition and measurement.
there are 2 requirements for developing and using predictors:
a - they must be relevant to the job
b - they must be appropriate ways to measure the employee attributes identified as critical to job success.
perdictors represent the measures of those employee attributes identified through a job analysis as being important for job success. Thus, a predictor should be used if there is a good reason to expect that the employee attributes assessed by the measure will predict the criteria that define job success. Tests, interviews, biographical data questionnaires, application blanks, and assessment centers are just some of the types of predictors.