H5. Assessment Methods for Selection and Placement. Industrial and organizational Psychology

If a biographical inventory contains enough of the second type of question, it begins to approximate a psychological test that assesses interests and personality instead of prior life experiences (Schmitt & Chan, 1998). Most inventories use a multiple-choice response format that can be easily scored. Possible answers to the question about enjoying college might be

"Enjoyed very much" "Enjoyed somewhat" "Enjoyed a little" "Didn't enjoy at all"

Group Versus Individually Administered Tests -A group test can be ...... -An individual test is one that ......

-A group test can be administered to several people at once. The test itself can be in a printed form (e.g., a booklet), or it can be displayed on a computer screen. In the former case, a group of people can be administered the test in a room at the same time, as is typically done with a course exam. In the latter case, several individuals can take the test simultaneously, either in the same location (e.g., a computer lab) or in different locations. In employment settings, there is often concern about testing conditions and test security, so a test administrator will control access to the test (e.g., hand out tests or log people onto a computer), time the test, and make sure that the individual taking the test is the candidate. -An individual test is one that a test administrator gives to a single test taker at a time rather than to a group of individuals. This is necessary because either the administrator has to score the items as the test proceeds or an apparatus is involved that only one person can use at a time. The test administrator sets the pace of the individual items. This approach is often used in administering cognitive ability tests to children. Because of its greater efficiency, the group test is preferred when it is feasible.

There are two types of interviews conducted in organizational settings. -During an unstructured interview , the interviewer ...... -By contrast, during a structured interview, the interviewer has .......

-During an unstructured interview , the interviewer asks whatever questions come to mind. It can be much like a conversation between the interviewer and interviewee in which the nature of the interaction between the two people determines in large part what is discussed. -By contrast, during a structured interview, the interviewer has a preplanned series of questions that are asked of every person who is interviewed. This makes the interview relatively standard, although the interaction between the two people can still affect what gets discussed. Nevertheless, the use of a standard set of questions allows the interviewer to collect the same information about each interviewee.

An interview has two advantages over an application form.

-First, the interview allows for longer and more detailed answers to questions that do not have short or simple answers. Most people find it easier to talk than write. -Second, the interview allows both the interviewer and the interviewee to ask one another for clarification. The interviewer can ask the interviewee to explain an unclear answer or to provide additional details. The interviewee can ask the interviewer to rephrase an unclear question or to indicate if the appropriate level of detail has been given.

Researchers frequently use personality tests to study many different aspects of people's behavior in organizations, but there are two major concerns with the use of personality tests for employee selection. -First, there is the possibility that job applicants will ...... -Second, personality tests do not always appear to be as ......

-First, there is the possibility that job applicants will fake responses to the test by answering the items in the manner they believe will put them in the most favorable light (Birkeland, Manson, Kisamore, Brannick, & Smith, 2006). It is feared that faking will undermine the validity of the test (Heggestad, Morrison, Reeve, & McCloy, 2006). Research has revealed, however, that attempts at distortion will not necessarily invalidate a personality test used for employee selection (J. Hogan, Barrett, & Hogan, 2007). This might be because an individual who knows how he or she should appear on a personality test will also know how he or she should behave at work to be effective. Thus the test might be assessing a person's self-presentation skills (Marcus, 2009). -Second, personality tests do not always appear to be as job relevant as other assessment devices. Job relevance means that what an assessment tool measures is linked to the KSAO requirements of the job. A typing test, for example, is obviously relevant for a secretary who must type as part of the job requirements. It is harder to show that a personality trait such as dominance is important for job performance. This has led some to question whether or not it is appropriate to use these tests.

Power Versus Speed Tests A power test gives the test taker almost unlimited time to complete the test. A speed test has a strict time limit. It is designed so that almost no one can finish all items in the allotted time. There are two ways in which the speed test is used....

-First, a speed test can contain challenging items that must be completed under time pressure. Some instructors use speed tests for classroom examinations under the presumption that the better-prepared students will be able to answer the questions more quickly than those that are less well prepared. The drawback to this use of speed tests is that the test taker is at a disadvantage if he or she is a slow reader or writer. -The second use is with a test that is designed to assess a person's speed in doing a particular task. A typing test is timed because its purpose is to assess a person's typing speed, as well as accuracy.

There are two major disadvantages (McBride, 1998).

-First, developing a computer testing system can be expensive and time consuming, although the increasing availability of off-the-shelf electronic assessments can reduce that cost. Hardware and software costs can be considerable, especially compared to the small cost of printed test booklets and pencils. The use of computer-scannable answer sheets, however, allows for cheap computer scoring of large numbers of tests, thus eliminating one advantage of computer administration. -Second, computerized tests are not necessarily equivalent to printed tests, especially for speed tests, which time how many correct items a person can do in a given interval. One such test is a clerical speed and accuracy test in which the test taker must compare two strings of letters and indicate if they are the same or not.

Electronic Administration of Psychological Tests Paper-and-pencil psychological tests can be adapted for administration electronically. The test taker can read each item on a computer screen or display and respond with the keyboard, mouse, or other interface device. There are two major advantages to this approach.

-First, the test can be scored automatically as soon as the last item is answered, which can speed up the selection process. -Second, a test can be put on a website, allowing access from almost anywhere in the world. An applicant doesn't have to come to a particular testing site to take the test. Of course, steps must be taken to monitor test taking so that cheating doesn't occur.

Sorts of tests

-Group Versus Individually Administered Tests -Closed-Ended Versus Open-Ended Tests -Paper-and-Pencil Versus Performance Tests -Power Versus Speed Tests

JOB-RELATED CHARACTERISTICS Many different characteristics of people, or KSAOs (knowledge, skill, ability, and other personal characteristics), are needed for a job. (See Chapter 3 for a more extensive discussion of KSAOs in the context of job analysis.) Knowledge refers to ..... A skill is ...... Ability is ...... Other personal characteristics are .....

-Knowledge refers to what the person knows about a job, such as legal knowledge for an attorney. -A skill is something that a person is able to do, such as program a computer or type. -Ability is the capability to learn something, such as the ability to learn to play a musical instrument or to speak a foreign language. -Other personal characteristics are every other human attribute not covered by the first three. Included are interests, personality, physical characteristics (such as height or eye color), and prior experience relevant for the job.

The interview can be used in two ways. -One is as a.... -The other way the interview can be used is to ......

-One is as an alternative to an application form or a written questionnaire to collect information. Questions can ask about attitudes ("Did you like your last job?"); job experiences ("Have you ever supervised anyone?"); personal background ("What was your major in college?"); and preferences ("Would you mind working weekends?"). -The other way the interview can be used is to make inferences about a person's suitability for a job based on both the person's answers to the questions and his or her behavior in the interview situation. The interview can be considered as a sample of interpersonal behavior, showing how well a person communicates and relates to the interviewer.

There are two types of EI assessments (Joseph & Newman, 2010).

-The trait type measures EI as a personality characteristic that is accessible to the person's awareness. These trait assessments look like a personality test in that people respond to each item concerning how well it describes them. -The ability type has a multiple-choice format with one right answer per question. It is based on the assumption that people cannot directly report their EI because it is not accessible to awareness. Rather, a person reveals his or her EI by choosing the correct answers on the test, just as he or she would indicate cognitive ability with a general intelligence test.

Several factors contribute to the validity of the structured interview. Huffcutt, Roth, and McDaniel (1996) conducted a meta-analysis showing that cognitive ability is related to interview outcomes and is a factor in their validity. Dipboye and Gaugler (1993) discussed six factors that are listed in Table 5.3. These factors involve

-the design of the questions asked, -the information available and not available to the interviewer, -the ratings made by the interviewer, and -the training given to the interviewer. P. E. Roth and Campion (1992) found good validity for a structured interview that contained most of the factors noted as important by Dipboye and Gaugler.

This chapter discusses five techniques for the assessment of characteristics that are frequently used for selection and placement. A psychological test consists of ...... Biographical information forms ask about...... The interview is ...... A work sample is ...... An assessment center is .....

A psychological test consists of a standard set of items or tasks that a person completes under controlled conditions. Most involve paper-and-pencil tasks, such as answering questions or solving problems, although some involve manipulation of physical objects to assess such characteristics as manual dexterity and eye-hand coordination. Psychological tests can be used to assess ability, interests, knowledge, personality, and skill. Biographical information forms ask about relevant prior experiences, such as level of education and work experience. Some forms can be quite detailed, asking not only about objective facts but about opinions and subjective reactions as well. The interview is a meeting between the job applicant and someone at the employing organization who will have input into the hiring decision. This can be done face to face or via technology such as a telephone or webcam. A work sample is a test that asks a person to perform a simulated job. The person is given the necessary materials and tools and must perform a particular task, such as assembling a motor, under controlled conditions. An assessment center is a series of exercises, including simulated job tasks, that measure how well a person can perform a job. It is commonly used to assess potential for a management or other white-collar job.

Multiple items provide increased reliability and validity over a single indicator of the characteristic. Single-item measures tend to have low reliability because a person can easily make a mistake on any one item. For example, an item can be misinterpreted or misread. Consider the following item that might be encountered on a test:

I'm not usually the first one to volunteer for a new work assignment.

The overt integrity test assesses a person's attitudes and prior behavior. It asks the person to indicate agreement or disagreement with statements concerning honesty and moral behavior. An example of such an item would be the following:

It is alright to lie if you know you won't get caught. The test taker is also asked how often he or she has done a number of counterproductive behaviors. A possible item might be the following: How often have you stolen something from your employer?

1 KSAOs for a Computer Sales Associate

Knowledge Knowledge of computer systems Knowledge of computer software Skills Skill in using a cash register Skill in completing monetary transactions Abilities Ability to understand complex technology Ability to communicate with other people Other Neat appearance Outgoing, friendly personality

TABLE 5.4 Dimensions Scored in an Assessment Center

Oral communication Oral comprehension Problem solving Interpersonal relations Coaching Planning Written communication

CHAPTER SUMMARY One of the major tasks I/O psychologists do for organizations is the assessment of people's characteristics for selection and placement. These characteristics can be classified as the knowledge, skill, ability, and other personal characteristics, or KSAOs, necessary for successful performance on the job. The five major methods used to assess KSAOs are:

Psychological test Biographical inventory Interview Work sample Assessment center

Campion, Palmer, and Campion (1997) discussed 15 ways in which structure can be introduced into an interview. There is no one way in which such interviews are conducted. Some interviews are ..... Others have strict ...... In the extreme, the interview is much like an ....

Some interviews are moderately structured in that the questions to be asked or just the topics to be covered are specified and standardized. Others have strict requirements that questions be asked in a set order, that the same phrasing be used, that interviewees not be asked to elaborate on any answer even if it is unclear, and that interviewees not be allowed to ask questions themselves until the end. In the extreme, the interview is much like an open-ended test with an oral rather than a written response. Campion et al. suggested that the more highly structured approaches are probably the most valid, and therefore as much structure as possible should be introduced into the interview.

One of the most popular vocational interest tests is the Self-Directed Search (Holland, 1994). This test provides scores on six personality types (Figure 5.7). Each type is associated with a particular family of occupations. As shown in the figure, the Investigative type likes investigative-type jobs that include scientific fields such as biology and geology. The profile of scores on the six types can guide a person in choosing a career.

The Realistic type likes jobs such as automobile mechanic, aircraft controller, surveyor, farmer, or electrician. The Investigative type likes investigative jobs such as biologist, chemist, physicist, anthropologist, geologist, or medical technologist. The Artistic type likes artistic jobs such as composer, musician, stage director, writer, interior decorator, or actor/actress. The Social type likes social jobs such as teacher, religious worker, counselor, clinical psychologist, psychiatric case worker, or speech therapist. The Enterprising type likes enterprising jobs such as salesperson, manager, business executive, television producer, sports promoter, or buyer. The Conventional type likes conventional jobs such as bookkeeper, stenographer, financial analyst, banker, cost estimator, or tax expert.

All test takers are not proficient in the local language; therefore, some tests have been developed that do not rely on reading ability. Figure 5.2 contains two items from the Beta II, which is a nonverbal intelligence test. The items involve problem solving without words.

The first item asks the test taker to find his or her way through the maze. The second item asks the test taker to solve a coding problem using numbers and shapes. The test administrator reads instructions to test takers for each type of item. Instructions can be given in any language.

Factors That Can Enhance the Reliability and Validity of a Structured Interview

The interviewer should ask standardized questions. The interviewer should have detailed information about the job in question. The interviewer should not have prior information about the interviewee, such as psychological test scores. The interviewer should not evaluate the interviewee until the interview is completed. The interviewer should make ratings of individual dimensions of the interviewee, such as educational background and relevance of prior work history, rather than a global rating of suitability for the job. Interviewers should receive training in how to conduct a valid structured interview.

Research on integrity tests has shown that they can predict counterproductive behavior at work, including absence, theft, and other behaviors potentially damaging to organizations (Ashton, 1998; Ones & Viswesvaran, 1998). This is true for both overt and personality integrity tests, which are also related to one another (Marcus, Lee, & Ashton, 2007). Interestingly, research has shown that integrity tests may do

a better job of predicting absence, general counterproductive behavior, and job performance than they do theft itself (Wanek, 1999). Part of the problem with conducting validation studies is that accurate data on employee theft are difficult to obtain because many employees who steal are never caught. Thus it is difficult to determine how well the test can predict stealing when you don't know for certain which employees were dishonest. However, as Wanek pointed out, these tests have been shown to predict useful behaviors at work besides theft, so there can be benefits beyond control of theft from using integrity tests for selection. For example, Van Iddekinge, Taylor, and Eidson (2005) showed that a commercially available integrity test was able to predict job performance.

In a leaderless group exercise, several assessees are given a problem to solve together. The problem might be

a competitive one, such as dividing up a scarce resource such as deciding which department gets a new piece of equipment. Each member of the group might be asked to role-play a particular management position. The problem can also be a cooperative one, in which all group members must generate a solution to an organizational problem, such as deciding whether or not to market a new product.

Work samples have been found to be good predictors of future job performance (Robertson & Kandola, 1982). The close correspondence between the assessment situation and the job itself certainly has much to do with the success of the work sample in predicting performance. As with most simulation techniques, the work sample has

a high degree of job relevance. This increases the likelihood that people will accept its use as a predictor of important skills in employee selection. There are several limitations to keep in mind, however. For many work samples, the applicant must already have experience with the task, work samples can be costly to develop, and the work sample is specific to a particular type of job (Callinan & Robertson, 2000).

Assessment center exercises have a high level of realism because they simulate many of the actual tasks included in the job. For example, the person being assessed, or assessee, may be asked to role-play being a manager in a given situation. This might involve dealing with a subordinate or handling the paperwork for the job. The assessee's performance on each exercise is scored by

a panel of trained assessors. The assessors are asked to evaluate the person's performance on each of several dimensions relevant to the job in question. For a manager's job, these can involve communicating, dealing with other people, making decisions, and planning. Table 5.4 contains the dimensions from a typical assessment center for managers described by Harris, Becker, and Smith (1993).

The problem with assessment centers is that the various dimension scores given to assessees within an exercise, such as an in-basket exercise, are too highly correlated. Furthermore, the scores given to assessees on the same dimensions across different exercises are not correlated enough. The high correlations across dimensions within exercises suggest that assessors are assessing only

a single dimension rather than multiple dimensions. One possibility is that assessors are able to judge only overall exercise performance rather than the individual dimensions. The low correlations between dimensions across exercises suggest that each exercise might assess a different characteristic.

Perhaps the most widely used performance test is

a typing test. With this sort of performance test, the test taker demonstrates his or her typing ability on an actual keyboard under standardized conditions. This tests the typing ability itself rather than typing knowledge, which could be assessed with a paper-and-pencil test of knowledge about typing.

One such test is a clerical speed and accuracy test in which the test taker must compare two strings of letters and indicate if they are the same or not—for example:

abdiel versus abdifl ghicbe versus ghicbe Response time per item can be different when using computer versus paper and pencil for the same items, making scores nonequivalent.

Psychomotor Ability Tests Psychomotor ability tests assess such things as

ability to manipulate objects and use tools. They involve both the coordination between senses and movement (e.g., eye-hand coordination) and the accuracy of movements. Many psychomotor tests are performance tests rather than paper-and-pencil tests because the abilities of interest involve manipulation of objects rather than cognitive elements. People are scored on their ability to perform motor tasks, such as putting pegs in holes or using simple tools to manipulate objects.

If a person misreads the item and does not notice that the second word is not , the meaning of the item will be reversed. His or her response to the item will be the opposite of what it should be, and the item will not be

an accurate indicator of the characteristic of interest. If it is likely that only a few people will make this error, the item may retain some reliability and validity. The reliability and validity are better for a multiple-item test than that for a single-item test because the impact of occasional errors on each person's score is reduced. The contribution of each item to the total score when there are many items is quite small.

Knowledge and Skill Tests An ability test is intended to assess a person's capability or potential to learn. A knowledge and skill test, often called an achievement test, is designed to

assess a person's present level of proficiency. A knowledge test assesses what the person knows, and a skill test assesses what a person is able to do.

Means to improve assessment center ratings were also found by Arthur, Woehr, and Maldegen (2000). They were able to produce good correlations within dimensions across exercises by carefully constructing their assessment center. Most of their assessors were I/O psychologists, and all received

careful training about how to translate specific behaviors into ratings. Finally, Lievens (2001) was able to increase the accuracy of assessment center ratings through the use of frame-of-reference training (see Chapter 4). Raters were given examples of effective and ineffective behaviors so they could build a common understanding of performance during training. A combination of all these approaches—using checklists, offering frame-of-reference training, and more carefully choosing assessors and conducting the centers—will likely produce the best results.

Integrity Tests An integrity test is designed to predict whether or not an employee will engage in counterproductive or dishonest behavior on the job. The behaviors these tests are designed to predict include

cheating, sabotage, theft, and unethical behavior. They also are sometimes used to predict absence and turnover.

Although the overall scores in an assessment center have been shown to be valid, questions have been raised about the construct validity of the individual dimension scores (Bowler & Woehr, 2009; Lance, Lambert, Gewin, Lievens, & Conway, 2004). In other words, we are not certain that the intended dimensions of the assessment center are actually what is being measured. Different exercises are designed to assess a

common set of dimensions. Because the same dimensions are assessed across exercises, scores on corresponding dimensions from different exercises should correlate well. For example, the score for leadership in the in-basket exercise should relate to the score for leadership in the leaderless group exercise. On the other hand, the different dimension scores within the same exercise should not be highly correlated because the dimensions are supposed to be measures of different, distinct characteristics. High correlations among dimension ratings are similar to halo errors in performance appraisal (see Chapter 4).

All five of the assessment techniques that we discuss attempt to measure KSAOs that are relevant for job performance and other organizationally relevant variables. As with all assessment techniques, the properties of reliability and validity are critical (see Chapter 2 for a discussion of these properties). That is, all measures must be

consistent (reliable) and must pass stringent tests for validity. In other words, evidence must exist that they can accomplish the tasks for which they are used in organizations. If a test is to be used to select police officers, for example, it must be shown to predict how well a police officer will do on the job.

Research on the assessment center has found it to be a valid device for the selection of employees (Arthur, Day, McNelly, & Edens, 2003). This means that scores on the assessment center are

correlated with job performance. For example, Shechtman (1992) found that scores in an assessment center given to university students upon admission predicted their job performance as teachers from 2 to 5 years after graduation, with a correlation of .27. Dayan, Kasten, and Fox (2002) found that assessment center ratings predicted job performance both 2 and 4 years later for police officers, with correlations of .30 and .21, respectively.

Research efforts have been made to improve the validity of dimension ratings in assessment centers. Reilly, Henry, and Smither (1990) reasoned that perhaps assessors are unable to adequately assess individual dimensions because they have too much information to process. Each exercise provides a great deal of information, and often that information must be organized in some way for judgments to be made. To help assessors organize information, the researchers provided a checklist of 273 behaviors to use to rate performance in the exercises. The assessors used the checklist to note the specific behaviors performed by each assessee before making their ratings. The results of the study showed that

correlations between dimensions within exercises got smaller and correlations between corresponding dimensions across exercises got larger than those typ- ically found in assessment centers. Thus the validities of the dimension scores improved. These results suggest that one problem with assessment centers as they are currently conducted is that they overload the assessor's ability to accurately assess dimensions. Providing checklists of behaviors imposed structure on the judgment task and improved the validity of the exercise ratings.

WORK SAMPLES A work sample is an assessment device that requires a person to

demonstrate how well he or she can perform the tasks involved in a job under standardized conditions. It is a type of simulation in which a person does a job or part of a job under testing conditions rather than actual job conditions.

The Bennett Mechanical Comprehension Test. This test assesses a combination of mechanical ability and knowledge about tools. This sort of test is useful for ...

determining if a person has a good understanding of how mechanical things work and how tools can be used. Note that this is a knowledge test and assesses a person's knowledge about tools rather than his or her skill in using tools. Actual tool use is a skill that involves both cognitive and psychomotor elements.

The disadvantage of the interview is that the interviewer can affect the answers of the interviewee. Even with the most carefully constructed structured interview and thoroughly trained interviewer, the interaction between interviewer and interviewee will

differ from interview to interview. Differences among interviewers in their ability to conduct an interview and in their personalities could affect the interview process and the interviewee's responses (Dipboye & Gaugler, 1993). This is not a concern with applica- tion forms because there is relatively little interaction between the person administering it and the person completing it.

An in-basket exercise asks the assessee to pretend that it is the first day of a new job and he or she has found a series of items in his or her in-basket. Items include

e-mails, letters, memos, and phone messages. The assessee's task is to deal with each item in an appropriate manner, deciding what action, if any, to take by making notes on each item.

BIOGRAPHICAL INFORMATION One of the easiest ways to find out about people is to ask them what you wish to know. In an employment setting, basic information about people is obtained from an application form. Although they can differ from organization to organization, standard application forms contain questions about

education, job skills, personal characteristics, and work history. Some forms can be quite detailed, asking about specific experiences, such as extracurricular school activities (e.g., participation in sports). Most ask about biographical information concerning education and relevant work experience that might be necessary for a particular job.

-An individual test is one that a test administrator gives to a single test taker at a time rather than to a group of individuals. This is necessary because

either the administrator has to score the items as the test proceeds or an apparatus is involved that only one person can use at a time. The test administrator sets the pace of the individual items. This approach is often used in administering cognitive ability tests to children. Because of its greater efficiency, the group test is preferred when it is feasible.

Stokes and Searcy (1999) distinguished

empirical from rational biographical inventories. -The empirical biographical inventory is developed by administering a large number of potential items to a group of employees in a particular job. Those items that are related to job performance are retained for the scale. -The rational biographical inventory is developed by beginning with an analysis of KSAO requirements and then devising items that reflect those KSAOs.

Each of the five assessment techniques can be used to determine a person's suitability for a particular job. Often more than one technique is used at the same time to get a more complete picture of how well an individual's characteristics match those necessary for a job. These assessment techniques can have uses other than selection and placement, however. They can be useful for

employee development by showing a person's strengths and weaknesses, which can then be addressed by training. Many of these techniques are used to assess human characteristics in research studies concerned with almost all the topics discussed throughout this book. In the next chapter, we will see how the five techniques are used to help choose people for jobs.

the Big Five personality theory dimensions of

extraversion, emotional stability, agreeableness, conscientiousness, and openness to experience.

Many different knowledge and skill areas can be assessed with a test. Some tests focus on

general skills, such as math and reading. Others are useful for assessing skills at particular job tasks, such as typing. With a typing test, the test taker is given the task of typing some materials under standardized conditions. The person is scored on the number of words per minute typed and the number of typing errors. As with ability tests, some knowledge and skill tests are paper-and-pencil tests and some are performance tests.

Assessors rate each assessee on each dimension by both observing behavior and reviewing materials produced during the exercises. Each person is scored on the various dimensions and may get an overall score. The dimension scores can be used to

give the person feedback about his or her strengths and weaknesses. The overall score is useful if assessment center performance is to be used to make hiring or promotion decisions.

Computer Adaptive Testing Computer adaptive testing (CAT) is a flexible computerized approach to item administration where items given to a test taker are chosen based on prior correct or incorrect responses. For an ability test where some items are more difficult than others, the session begins by

giving the test taker a moderately difficult item. If the person gets the item correct, a more difficult item is presented. If the person gets it incorrect, an easier item is chosen. As the test proceeds, the computer chooses items that are at the appropriate level of difficulty for the particular test taker. Thus each test taker gets a somewhat different set of items, depending on his or her pattern of correct and incorrect answers. By doing this, CAT is more efficient than standard testing because it can achieve the same level of reliability with fewer items (Penfield, 2006).

The personality integrity test assesses personality characteristics that have been found to predict counterproductive behavior. Whereas the overt integrity tests are obvious assessments of honesty and integrity, the personality tests are

hidden in their purpose. In fact, standard personality inventories often are used as integrity tests.

It might seem that computer administration has the potential for increasing the accuracy of personality tests in which people often distort or fake their responses—people are sometimes better able to reveal themselves honestly to a machine, such as a computer, than to another person. A computer might seem more

impersonal than a paper-and-pencil test to a test taker.

A still more complex type of computer assessment that measures complex problem solving has become popular in Germany (Funke, 1998; Kleinmann & Strauss, 1998). These assessments provide the applicant with a simulation of a complex work situation, such as running an airport or factory, that is somewhat like a computer game. The person is given a fixed amount of time to run the simulation, and various indicators of performance are calculated by the computer. Research with these assessments has provided

impressive initial evidence that they can predict future job performance for jobs in which problem-solving skill is important (Kleinmann & Strauss, 1998).

An assessment center can contain many different activities and exercises. The assessee might participate in an interview and take a battery of psychological tests, in addition to completing several simulation exercises. The exercises might include an

in-basket exercise, a leaderless group exercise, a problem-solving simulation, and role plays. Each exercise yields scores on several dimensions, and usually each dimension is assessed by several exercises. In addition, there is an overall assessment of the person's potential for the job in question.

Vocational Interest Tests A vocational interest test matches either the interests or the personality of the test taker to those of people in a variety of different occupations and occupational categories. Interests are assessed by asking the test taker to

indicate his or her preferences for engaging in various activities, such as between attending a sporting event and visiting a museum. Personality is relevant, as specific traits might drive people's interests; for example, people who are outgoing and extraverted might prefer activities that involve contact with other people (Larson, Rottinghaus, & Borgen, 2002). Data from vocational interest tests are available about the answers of people in many different occupations. The test taker's answers are matched to those of people in different occupations to see how well he or she fits each occupation.

Research with various EI tests has shown that they can predict job performance. For example, a meta-analysis by Joseph and Newman (2010) found that on average EI trait tests correlate .32 and EI ability tests correlate .16 with job performance. EI has also been shown to relate to students' college grade point averages (GPAs) (Rode, Mooney, Arthaud-Day, Near, Baldwin, Rubin, et al., 2007) and for managers to the job satisfaction of their subordinates (Sy, Tram, & O'Hara, 2006), although correlations are often quite small. However, research does not support the claims that EI

is one of the most important characteristics for job success. Joseph and Newman (2010) found that EI ability measures related to performance only in jobs that require employees to manage their own emotions—that is, to engage in emotional labor (see Chapter 9).

INTERVIEWS An interview is a face-to-face meeting between one or more interviewers who are collecting information or making hiring decisions and an interviewee. Almost all organizations use the interview in hiring for almost all positions. Perhaps one reason for the universal use of this technique is that

it is widely accepted. Studies across Europe and the United States have shown that people rate interviews as among the most fair selection procedures that organizations use (Bertolino & Steiner, 2007; Ispas, Ilie, Iliescu, Johnson, & Harris, 2010).

The KSAOs for each job can be determined by a detailed and thorough study, which is called a

job analysis . (See Chapter 3 for a discussion of job analysis.) This involves a number of techniques that result in a list of the necessary KSAOs for the job in question. Once KSAOs are determined, procedures can be chosen or developed to assess them in job applicants or current employees. The idea is to select or place people who have the necessary KSAOs for the jobs available. Although this process does not guarantee that all the people chosen will be successful on the job, it increases the chances of making good choices over using other selection and placement approaches.

Cognitive Ability Tests An intelligence or IQ test of general cognitive ability is the best-known cognitive ability test. There are also tests of individual cognitive abilities, such as mathematical or verbal ability. The items on such tests represent individual problems to solve. Such tests can be administered to

large groups of individuals simultaneously and provide an inexpensive and efficient means of assessing job applicants.

ASSESSMENT CENTERS An assessment center measures how well a person is able to perform some of the tasks of a specific job. It consists of several exercises that are designed to simulate various job tasks. Most assessment centers are designed to assess

management skills, but they are used for nonmanagement jobs as well. The exercises can take several days to complete and are often administered to several individuals at a time. This technique is widely used by many types of employers. For example, in a survey of British employers, Keenan (1995) found that 44% used assessment centers to hire college graduates. In the United States, the assessment center is used primarily for hiring and promotion decisions, although it can also be used to help employees enhance job skills (Bell & Arthur, 2008; Spychalski, Quin˜ones, Gaugler, & Pohley, 1997).

PSYCHOLOGICAL TESTS A psychological test is a standardized series of problems or questions that assess a particular individual characteristic. Psychological tests are commonly used to assess ....

many KSAOs, including knowledge, skills, abilities, attitudes, interests, and personality. They are comprised of multiple items, which are indicators of the characteristic of interest. Each item can be completed relatively quickly, making it feasible to include many items to assess each characteristic and to assess several characteristics at one time.

There is also less concern about security than with nonadaptive testing, since each test taker gets a different set of items, thus making it more difficult to cheat (Reynolds & Dickter, 2010). For these reasons, this approach has been used for

masstesting, such as the computerized version of the Graduate Record Examination (GRE). CAT is not without disadvantages, however. Perhaps the biggest is that the tests take considerable time and resources to develop. A well-designed test will likely have hundreds of items, and data on each item must be collected from a thousand or more test takers in order to determine difficulty levels (Reynolds & Dickter, 2010). Thus it takes a lot more time and money to develop a computer adaptive test than a nonadaptive test.

Interest and personality profiles vary considerably from occupation to occupation. Therefore, any individual test taker will match some occupations and not others. Because occupations tend to group together into categories, a person can be

matched to an occu- pational category. One such occupational category is artistic, which involves creative processes, such as interior design and photography.

An important issue with EI research, however, is that some have questioned the construct validity of the EI measures used in these studies, suggesting that EI ability measures overlap with

measures of cognitive ability and personality (Christiansen, Janovics, & Siers, 2010), which are the real underlying factors relating to performance (Landy, 2005). At the current time, we know that EI measures relate to a number of important variables, but the construct validity of these measures is in dispute among researchers.

The structured interview can be a good way to collect information that builds on that provided in the application form. Specific questions can ask for

more detail about the provided information. For example, an application form will typically ask the person to list all prior jobs. In the interview, the interviewee might be asked to describe each job and to explain how his or her prior work experiences might be relevant to the job at hand.

Richman, Kiesler, Weisband, and Drasgow (1999) conducted a meta- analysis to see if the tendency to respond in a favorable direction was the same for computer as for paper-and-pencil administration of personality tests. Overall, the results suggested

no differences. However, in looking more closely at the results, Richman et al. noticed that there was somewhat less distortion when the person was alone. When an experimenter was in the room, any feeling of anonymity was apparently lost, and there were no differences between the computer and paper-and-pencil administrations.

One of the limitations of the telephone is

not allowing for observation of nonverbal information—such as appearance, facial expression, and gestures—and these elements can be important to assess for some jobs such as sales. Furthermore, as Chapman, Uggerslev, and Webster (2003) found, applicants may have better attitudes toward face-to-face interviews than telephone or videophone interviews. Thus it seems unlikely that face-to-face interviews will disappear anytime soon.

Wanek, Sackett, and Ones (2003) noted that there are two different types of integrity tests—

overt and personality.

Paper-and-Pencil Versus Performance Tests With a paper-and-pencil test, the test is on a piece of paper or other printed (or electronic) medium, and the responses are made in written form, often with a pencil. A multiple-choice course examination is a paper-and-pencil test that presents the exam questions on paper, and the responses are made in pencil on the exam itself or on a separate answer sheet. Although we continue to use the name, quite often employment tests today are administered electronically, with responses made via keyboard or mouse. Open- ended tests can also be

paper-and-pencil tests if they ask people to record their responses in some form, either by writing on paper or by typing on a computer. A performance test involves the manipulation of apparatuses, equipment, materials, or tools.

Interviewers can be asked to make ratings on job-related dimensions, such as communication skills and relevant experience. They sometimes make overall ratings of applicant suitability for the job. These ratings can be subject to the same problems as

performance appraisal ratings, which we discussed in Chapter 4. Interviewer biases and cognitive processes can reduce the accuracy of judgments and ratings (Dipboye & Gaugler, 1993).

Despite concerns over the use of personality tests for employee selection, such tests have been growing in popularity among practitioners and researchers in organizations (Ones & Anderson, 2002). Several researchers have conducted meta-analyses to investigate the relation between personality and job performance. This technique of data analysis allows for the statistical combination of results across multiple studies of the same phenomenon (see Chapter 2). Such analyses have shown that personality can predict

performance, although the magnitude of these correlations is modest (Dudley, Orvis, Lebiecki, & Cortina, 2006; Salgado, 2003).

Ability Tests An ability or aptitude is a

person's capacity to do or learn to do a particular task. Cognitive abilities, such as intelligence, are relevant to tasks that involve information processing and learning. Psychomotor abilities, such as manual dexterity, involve body movements and manipulation of objects. The importance of each ability is dependent on the nature of the tasks of interest. Some job tasks require mostly cognitive abilities (e.g., programming a computer), whereas others rely mainly on psychomotor abilities (e.g., sweeping a floor). Many tasks require both types of abilities (e.g., repairing a computer or complex piece of equipment).

Biographical inventories been shown to predict job performance (Breaugh, 2009; Stokes, Toth, Searcy, Stroupe, & Carter, 1999). In a direct comparison, Stokes and Searcy (1999) found similar predictive power for both the empirical and the rational types of inventories. Furthermore, biographical inventories have been found to add predictability over and above

personality tests, suggesting they don't just assess the same thing (McManus & Kelly, 1999 —see the Research in Detail box; Mount, Witt, & Barrick, 2000). Furthermore, biographical inventories tend to be resistant to applicant faking in order to look good on the test because many items are factual in nature and give the appearance of verifiability (J. West & Karas, 1999).

With all the research focus on the validity of interviews, issues about what interviews actually assess have been overlooked. Huffcutt, Conway, Roth, and Stone (2001) did an extensive analysis of the specific personal characteristics assessed in interviews that provides insight into why they are able to predict future performance. The authors conducted a literature review and identified 47 interview studies that provided information about the KSAOs covered. The most commonly assessed were ....

personality traits and interpersonal skills. Furthermore, structured and unstructured interviews tended to focus on different things. Structured interviews targeted job knowledge and skills, person-organization fit, and interpersonal skills, whereas unstructured interviews targeted general intelligence, education, and prior experience.

Technology has also made it possible to conduct interviews remotely. Telephone and webcam interviews have become increasingly popular when applicants are located at great distances. Research has shown that this can result in higher ratings for interviewees in telephone interviews (Straus, Miles, & Levesque, 2001) and videophone interviews (Chapman & Rowe, 2001). This was particularly true for individuals who were

physically unattractive (Straus et al., 2001), suggesting a possible bias in interviews where the interviewer can see the interviewee. Schmidt and Rader (1999) conducted a meta-analysis of structured telephone interview studies and found that the average correlation with performance was about the same as for face-to-face interviews.

Computers are not the only technology that is being used to increase the efficiency of assessment. Van Iddekinge, Eidson, Kudisch, and Goldblatt (2003) developed a telephone- based system for administering a biographical inventory. Applicants were able to dial a toll-free number and answer yes/no questions about their backgrounds using the keys on their telephones. Van Iddekinge et al. showed that this type of assessment was able to

predict job performance for those subsequently hired. However, it required designing an inventory in a yes/no format, which is more limiting than other media, such as pencil and paper.

Drug Testing Organizations in the United States and elsewhere have turned to drug testing with both applicants and employees as a means of controlling drug use at work, and estimates are that as many as 30 million Americans undergo such tests each year (Mastrangelo & Popovich, 2000). Such testing is of particular importance in jobs that are

safety- sensitive, meaning that impaired performance could lead to accidents or injury, such as air traffic controllers and bus drivers. However, such tests have been controversial, with many people considering them an unfair invasion of privacy. Paronto, Truxillo, Bauer, and Leo (2002) surveyed 1,484 individuals about their attitudes toward drug testing. They found that respondents considered drug testing more acceptable for jobs that were safety-sensitive than for jobs that were not.

Closed-Ended Versus Open-Ended Tests With a closed-ended test, the test taker must choose one from several possible responses, as in a multiple-choice test. An open-ended test is like an essay exam. The test taker must generate a response rather than choosing a correct response. Whereas the closed-ended test is preferred because of its greater ease in scoring, the open-ended test is more appropriate for

some characteristics. For example, writing ability is best assessed by asking a person to write an essay. Experts can read and score the essay for a number of characteristics, such as clarity of expression and grammatical accuracy. These characteristics would be more difficult, if not impossible, to assess with a closed-ended test.

The biographical inventory asks much more detailed background questions than a typical application form. Whereas application forms ask about prior level of education and work experience, the biographical inventory asks about

specific experiences at school and work, or even in other areas of life. Some of the questions ask about objective, verifiable facts, such as "What was your grade point average in college?" Others ask about opinions or subjective experiences, such as "Did you enjoy college?"

In a problem-solving simulation, the assessee is given a problem and is asked to come up with a solution, perhaps by producing a report. The problem provides

sufficient background information from which to write the report. For example, the assessee could be given information about the costs for and projected income from opening a new assembly plant. His or her task would be to produce a feasibility report for opening the plant.

-A group test can be administered to several people at once. The test itself can be in a printed form (e.g., a booklet), or it can be displayed on a computer screen. In the former case, a group of people can be administered the test in a room at the same time, as is typically done with a course exam. In the latter case, several individuals can take the test simultaneously, either in the same location (e.g., a computer lab) or in different locations. In employment settings, there is often concern about

testing conditions and test security, so a test administrator will control access to the test (e.g., hand out tests or log people onto a computer), time the test, and make sure that the individual taking the test is the candidate.

Emotional Intelligence Tests A characteristic that falls between a personality trait and a cognitive ability is emotional intelligence (EI). It is defined as

the ability people have to control and recognize emotions in themselves and in others. In theory, this ability makes people more socially skilled, enabling them to be aware of and to control their impact on others. In a work setting, this means being able to work smoothly with colleagues, and in supervisory positions, it means the ability to exert leadership. Salovey and Mayer (1989) are credited with kicking off interest among psychologists in the idea that emotional intelligence is an important ability. Unfortunately various researchers who study it have adopted somewhat different definitions and conceptions, and exaggerated claims both inside the field and in the popular media have caused confusion and controversy.

One of the earliest applications of psychology to the human problems of organizations had to do with

the assessment of people for selection and placement. During World War I, the U.S. Army became one of the first organizations to use large-scale testing of people to determine their job assignments (placement). After the war, large organizations saw the potential value of assessing job applicants for selection and other employment decisions, and the use of testing and other techniques became commonplace. This is true not only in the United States but also in most of the industrialized world, including Canada, Western Europe, and Israel (McCulloch, 1993).

The typical work sample gives the applicant the materials and tools necessary to accomplish the task. The person is instructed to complete the task quickly but accurately. A score is computed based on the accuracy with which the task is completed and the amount of time it took. For example, a person might be asked to disassemble and reassemble an electric motor or a small gasoline engine. A trained observer would score the person on accuracy and speed. Perhaps the most familiar work sample is

the driving test that is required of applicants for a driver's license. The applicant is asked to complete a series of maneuvers with the automobile, while the test administrator records the scores for each one. To get the license, the applicant must achieve a total score that meets a prearranged criterion. A work sample in an organizational setting might be used the same way to determine if a person is suitable for a particular job.

In practice, it is difficult to totally separate ability from knowledge and skill, for ability tests rely to some extent on knowledge and skill, and knowledge and skill tests rely to some extent on ability. The major difference between the two types of tests is

the emphasis placed on prior knowledge and skill in performing specific tasks. For example, a psychomotor ability test might assess how quickly a person could put pegs in holes as an indicator of eye-hand coordination, whereas a psychomotor skill test might assess how well a person can type. Typing skill is the product of several different abilities.

Personality Tests A personality trait is

the predisposition or tendency to behave in a particular way across different situations. A person who prefers to do things with other people is said to be high on the trait of sociability. A person who frequently influences others is said to be high on the trait of dominance. Personality traits can be important because certain classes of behavior can be relevant for job performance and other behaviors in organizations. Sociability can be an important trait for a salesperson who must interact with other people, whereas dominance can be an important trait for a supervisor who must direct the activities of others.

A work sample is like a psychological test except that it is designed to measure a higher-level skill. A test measures a basic skill, such as eye-hand coordination or manual dexterity. A work sample assesses

the skill in doing a particular task, such as driving a bulldozer, which is composed of several basic skills performed in the context of a particular set of conditions. For some applications, the higher-level skill may be more important to assess than the basic skills because a work sample indicates how well a person can actually do a particular task. The psychological test indicates if a person has the requisite basic skills that should in theory predict how well he or she could do the task—but not how well he or she can actually do it.

Research evidence supports the use of structured interviews, but not necessarily unstructured ones, in making employment decisions. Ratings of interviewee employment suitability from a structured interview have been shown to predict future job performance across many different studies (Judge, Higgins, & Cable, 2000; Wiesner & Cronshaw, 1988). Wiesner and Cronshaw (1988) conducted a meta-analysis of over 100 interview studies in which they compared structured and unstructured interviews. The average correlation between interview outcome and job performance was larger for

the structured interview studies (r = .34) than for the unstructured interview studies (r = .17). Schmidt and Zimmerman (2004) argued that much of the difference in interview validity can be attributed to the lower reliability of unstructured interviews compared to that of structured interviews.

ELECTRONIC ASSESSMENT One of the fastest-growing trends in organizational assessment is the use of electronic media rather than paper and pencil. What began as a move to computerize test administration and scoring on personal computers has exploded with new kinds of assessments and approaches being developed. Lievens and Sackett (2006), for example, discussed the use of video-based assessment, where each item is presented as a video clip illustrating an event that might occur at work. Such assessments can provide a realistic situation that is difficult to describe fully with text. One of the major catalysts for electronic assessment has been

the web, which makes it possible to link an applicant with an assessment almost anywhere in the world. Applicant recruitment services, such as Monster.com, have incorporated assessments into what previously was the initial application process. Electronic technology is expanding the use of assessment, as it has made that assessment cheaper and more convenient.

One difficulty with drawing conclusions about the validity of personality tests from the existing meta-analyses is that

those analyses group together tests that assessed many different traits for the prediction of performance on many different types of jobs. Some of those traits may have represented job-relevant characteristics; others may have been unimportant for success. Existing evidence suggests that specific personality traits chosen to be relevant to job performance on particular jobs will do an even better job of predicting performance (Paunonen, Rothstein, & Jackson, 1999; Tett, Steele, & Beauregard, 2003).

The biggest criticism of biographical inventories has been limited to the empirical variety and concerns the fact that items are chosen based entirely on their ability to predict performance rather than on an apparent link to job requirements. Sometimes an item can predict performance even though it cannot be linked to a KSAO necessary for job performance. These items are combined into scores that predict performance but do not reflect job-related KSAOs. Many questions in such inventories can appear completely unrelated to the job or job performance. Applicants may consider some questions....

to be an invasion of privacy, such as questions about high school dating behavior. Finally, such inventories can often have limited generalizability (Mumford, 1999) and might be useful only for the job and location for which they were developed. On the other hand, rational inventories can often be used for many jobs and locations.

The match between a person's vocational interests and the nature of his or her job is related to how satisfied the person is with that job (Rottinghaus, Hees, & Conrath, 2009). A person who takes a job that is a poor match will likely be unhappy with it, whereas a person who takes a job that is a good match will probably like it. The idea behind the tests is

to encourage people to select careers that match their interests. Vocational interest tests are frequently used for vocational guidance in helping people decide what careers to pursue.

A role-play exercise requires that the assessee pretend to be a particular person in a specific organizational role, such as the manager of human resources. The task is

to handle a problem or situation, such as counseling a troubled employee or dealing with an irate customer.

With unstructured interviews, multiple interviewers will not agree very well on the applicant's suitability to be hired, and this lack of agreement hurts the ability of the interview to predict future job success. Another problem with the unstructured interview is that it leaves

too much room for biases. An interviewer might decide on the basis of appearance, for example, that an applicant is unsuited. Cable and Judge (1997) found that interviewers' hiring decisions were related to how attractive they found the applicant and how much they liked him or her. Interviewers might not be as thorough in questioning a disliked or unattractive applicant and might fail to discover important job-related qualifications that might lead to hiring. Stewart, Dustin, Barrick, and Darnold (2008) conducted simulated interviews that showed how eye contact and firmness of handshake affect interviewers' ratings of interview performance. Given the existence of biases, it should not be surprising that interviewers fail to agree about applicants.

Paper-and-pencil personality tests can be useful tools for assessing personality traits. Some personality tests are designed to assess a single personality trait; personality inventories assess multiple dimensions and are sometimes used to provide profiles of individuals across several personality traits. They can also characterize people into

types, which are combinations of different traits. For example, the extraverted type of person tends to be high on the traits of activity, optimism, sociability, and talkativeness, whereas the introverted type tends to be high on carefulness, quietness, passivity, and unsociability (Pervin, 1993).

Research has consistently shown that cognitive ability tests are

valid predictors of job performance across a large number of different kinds of jobs (Ones, Dilchert, Viswes- varan, & Salgado, 2010). People who score well on cognitive ability tests tend to perform better on the job. Cognitive ability tests have a long history of use by large organizations for employee selection because of their efficiency and validity. In a survey of 703 SIOP members, Murphy, Cronin, and Tam (2003) found consensus among I/O psychologists that such tests are fair and useful for selection.

Potosky and Bobko (2004) contrasted web-based with paper-and-pencil tests. They noted that for speed tests, where the test taker has a limited amount of time, adjustments must be made for the load times of web pages. Furthermore, test-taking strategies are likely to

vary between the two formats. It is easy with paper-and-pencil tests to skip around, so that the test taker might choose to complete certain types of items first and other types last. Since web-based tests allow fewer items per screen than paper does per page, it can discourage skipping around, and thus items are more likely to be completed in order. Care must be taken in comparing scores of tests taken with the two formats.

On balance, research on assessment centers raises the possibility that each exercise measures a single trait rather than multiple dimensions and that the different exercises may assess different things (Lance, Foster, Gentry, & Thoresen, 2004). Lievens, Chasteen, Day, and Christiansen (2006) suggested that each exercise provides a different set of opportunities for the assessee to display behavior relevant to performance in a unique way. One might reflect leadership and another might tap problem solving. Thus it should be no surprise that dimension scores are not very high across exercises. At the present time, we are not certain exactly

what the different KSAOs assessed might be. Thus use of the assessment center to give specific feedback about individual KSAOs (e.g., decision making or leadership) is questionable. Russell and Domm (1995) provided evidence that ratings reflect how well the individual performs the role for the job being simulated rather than how well he or she does each dimension. Regardless of why it works, the assessment center does a good job of predicting future job performance. For that reason, it is a popular assessment device for selection and placement.

Characteristics of Tests Many different types of tests are available that can assess hundreds of individual characteristics. The nature of the characteristic of interest helps determine

which test is used. For example, a test to assess mathematical ability will most certainly be composed of math problems. A test of physical strength, on the other hand, will likely involve the lifting of heavy objects. A discussion of the four distinguishing characteristics of tests follows.