Consequences of testing
During their entire lifetime, people are being tested. Examples of tests are the Apgar test for measuring the health of infants, driving and school examinations in adolescence, developmental tests, etc. Tests are being administered across many diverse contexts, and during our lifetime we are likely to have participated in so many tests that by the time of retirement they have had a major impact on our life course. A thorough knowledge of tests is therefore necessary for everyone in the field of psychology. Someone who develops and evaluates tests within the fields of psychology or education is called a psychometrist. Personality and intelligence tests are currently the most essential tests in psychology.
Definition of a test
Tests can be very different in their purposes and format, but in general they share the following characteristics:
A test is a standardized procedure for identifying behavior and describing it by means of categories or scores. There are a number of defining characteristics of tests. Firstly, a test is standardized, which means that the procedures for undertaking it are the same in different settings and under different conditions. Secondly, a test is based on a sample of the behavior that is to be measured. The items within the test do not have to cover the entire behaviour being viewed, as long as they are relevant in a way that permits making inferences about the behaviour. It is important that the behavior in the sample represents the behavior that is predicted by the test. Thirdly, it must be possible to derive categories or scores from the test. A certain amount of measurement error must always be taken into account: X = T + e, where X is the observed score, T is the true score and e is the error. A test developer should try to make e as small as possible.
It should also be remembered that the abstract characteristic that is measured by a test does not represent a physical 'something' in the world. Fourthly, it is necessary to establish a standard with which scores of participants can be compared. This is done by means of a standardisation sample, whereby this sample must be representative of the population for which the test is intended. The standard indicates when people deviate. Finally, tests are intended to predict nontest behaviors. A test can therefore have more than one goal and goals can differ from the actual behaviour being measured by the test. To know whether the behavior is actually predicted by the test is done through validation research, which is mostly being conducted after the test has been released.
Distinctions in testing
The majority of tests are norm-referenced, whereby the score of each participant is interpreted in comparison with a relevant standardized sample. Other tests are criterion-referenced, where the aim is to determine where a participant stands with regard to clearly defined criteria. School exams are an example of the latter category, because students’ scores are classified into a predetermined grade system (instead of making a comparison with a reference group).
Another important distinction is that between assessment and testing. Assessment is a term used for more comprehensive research and refers to the entire process of collecting information about a person, on the basis of which something can be said about characteristics and behaviour can be predicted. Tests are therefore only one source of information for an entire assessment process.
Types of tests
Tests can be roughly divided into group tests, which can largely be taken with pen and paper and with several participants at the same time, and individual tests, which are taken one-on-one. The various categories of tests are discussed below. They occur in different forms (norm-referenced, criterion-referenced, individual tests and group tests).
Intelligence tests: the general intellectual ability of an individual is measured, based on the skills that are important in a particular culture. There are sub-scores, but usually the general score is being used. The test generally consists of a heterogeneous combination of items that measure different aspects of intelligence.
Aptitude tests: one or more specific aspects of competence are measured. This kind of testing is often used to predict success in a particular job or study.
Achievement tests: this measures the degree of learning, success, or achievement of an individual with regard to a specific task. The difference with the aptitude test is the purpose and content of the test. Aptitude tests measure the course of the performance of individuals, Achievement tests measure the abilities of someone at the test moment.
Creativity testing: the ability to develop new ideas, insights or creations is measured. For these tests one has to be able to think divergent: looking for different solutions for a complex problem. There are still doubts about whether creativity is a form of applied intelligence.
Personality tests: characteristics or behaviors are measured that determine the individuality and character of a person.
Interest inventories: the preference of an individual for certain activities or topics is measured, often to help with occupational choice.
Behavioral procedures: the antecedents and consequences of behavior are measured, or the frequency of the behaviour is objectively described
Neuropsychological tests: these are used to investigate persons with possible brain damage. For the most part, these are long and intensive one-on-one tests.
Different types of test use
There are 5 ways to use psychological tests:
Classification: assigning people to certain categories. This can be subdivided into placement (assigning to different programs based on skills), screening (short tests to identify persons with special needs or characteristics), certification (whereby obtaining a test yields certain privileges) and selection (whereby certification provides access to 'private' circles such as the university or an association).
Diagnosis and treatment planning: determining the nature and cause of abnormal behavior and classifying the behavior within an accepted diagnostic system. Diagnosis must be more than a label, but it must also take into account the underlying information. The diagnosis is also used in the planning of the possible treatment plan.
Self-knowledge: gaining more insight into yourself through a test.
Program evaluation: evaluating the success of certain social or educational programs.
Research: testing hypotheses by means of tests.
The goals of testing often overlap, making a clear distinction difficult. Many tests can also be conducted once and still used for multiple purposes.
There are several factors that can affect the reliability of a test. These factors are discussed above.
Standardized procedures in test administration
Non-standardized tests can significantly influence the results, rendering them unusable. In addition, they are not valid. In some cases, however, it is desirable, sometimes even necessary, to be flexible with regards to the test procedure. This is the case, for example, with participants with disabilities. Deviations from the standard procedure should, however, always be intentional and well though through.
Desirable procedures of test administration
For individual testing it is important that the examiner is familiar with the material, the instructions he/she must give and the way in which details and scores are noted. In addition, it is very important that all participants can understand the written and spoken instructions. Account must also be taken with any restrictions of the participant in, for example, hearing, vision, speech, or motor control.
For people with reduced hearing, it is important that the test leader is aware of this and responds well to it, so that the test results are not influenced. As with vision limitation, most adults report these by themselves, but children often do not mention any limitations. Finally, possible limitations in motor control or speech must be taken into account. This is important in tests that use time responses. Tests can be adapted slightly for people with disabilities without the validity or reliability of the test deteriorating. Sometimes there also are special forms of a test that are adjusted to a certain limitation.
There are also a number of important points for group testing that have to be taken into account by the examiner. For example, when testing with a time limit, it is important that enough time is available and that it is attended to closely. In addition, instructions must be read clearly and not too quickly and must not be paraphrased. Background noise should also be limited as much as possible. Moreover, it is important to clearly indicate whether guessing, if the participant does not know the answer, does have any consequences. Many tests have a built-in guess correction.
Influences of the examiner
It is important that the test leader ensures 'rapport': a good rapport with the participants creates a comfortable and motivating atmosphere. This increases the cooperativeness with the participant. Research shows conflicting results about the influence of race, experience and gender of the examiner on the results. In some unique cases this seems to be of influence.
Background and motivation of the examinee
Different aspects of the participant can influence the test results. Test anxiety refers to all behavioral reactions that come along with concerns about possible failure of a test. Research shows that test anxiety is both a cause and consequence of poor performance on tests. Especially in tests with time pressure, the results of participants with test anxiety can be strongly influenced.
In addition, it sometimes happens that participants create false results to get a certain test result. You should also take into account the motivation of the participant: An unmotivated participant can provide unreliable results.
Professional standards for testing
Usually tests are carried out responsibly, but there are of course exceptions where the irresponsible application or interpretation of a test can sometimes have disastrous consequences. That is why guidelines for responsible test use have been developed by professional organizations such as the American Psychological Association (APA). The responsibilities of test developers and test users are described below.
Responsibilities of test publishers
Test publishers must take into account various factors. Firstly, tests must meet all guidelines before they are issued. For example, it is mandatory to provide technical and user manuals with the test. Secondly, any marketing and advertising of the test must take place in an accurate and sincere manner. A test may only be published when the reliability and validity have been investigated. The test must state in what way the reliability and validity were investigated and what the results were. It should also be clear who can use the test and what qualifications a person must have for this. Often certain certifications are required for use.
Responsibilities of test users
The APA, among others, has published ethical guidelines and professional standards for test use to ensure the well-being of the participants and their surroundings. This includes, for example, the guideline that testing must always be to the benefit of the client. Confidentiality is also a duty of the test leader, although it is mandatory to report serious threats to the participants or others.
In addition, it is necessary that the examiner has the necessary expertise to take a test. Informed consent is another important condition. This means that all participants are informed in advance about the research and give their permission. Furthermore, account must be taken of what the standard of care is for a specific case, i.e. which method or test is most frequently used at that moment and is the most accepted.
For example, one must be careful with the use of outdated material. In addition, test results should be communicated correctly with the participant, giving effective and constructive feedback. This should not be beyond the boundaries of the tester's expertise. The psychological report written about the research should be direct and concrete.
This is important because the content of the report can have an impact on the life of the participant, for example when the report is requested by an employer. Finally, respect and recognition of individual differences is very important for test use.
Testing cultural and linguistic minorities
Psychological tests are mainly aimed at Western populations. It therefore cannot be assumed that existing tests are suitable for all population groups. Since the 1930s, there has been a rise in culture-sensitive testing, but the work is far from complete. Other cultures may have different norms, values or beliefs. This may cause them to look differently at a test or respond differently to the results.
The influence of cultural background on test results
Research shows that people from different cultural backgrounds complete and interpret tests in different ways. For example, indigenous peoples in the US show a different conception of time than the white middle class in America.
In addition, it appears that, for example, African Americans qualitatively respond in a different way to testing than Anglo-Americans; children of African American origin turned out to be less spontaneous in their answers. Similar differences are also visible in adults. In addition, testing may involve the danger of stereotyping, whereby the participants unconsciously confirm the negative stereotype that exists about their own group. This is also called stereotype threat. Test scores are not always the same for individuals, but are created within a social psychological field that is influenced by different cultural factors.
Unintended effects of high-stakes testing
Another effect that can play a role in testing is fraud. This is particularly the case with tests whose results have a lot of influence, for example when selecting for a job or study. Mass fraud occurs seldomly. Moreover, fraud with the help of parents or teachers also tends to occur.
Another aspect of fraud is described by the Lake Wobegon Effect, which refers to the fact that in many schools more than 50% of pupils have above average grades. This is mainly because our society places great emphasis on performance and the excellence of schools. Teachers help to cheat the pupils by, among other things, coaching them on key answers, changing answer forms or giving them more time on tests and exams.
It seems that the national trend towards performance tests for selection and evaluation helps to encourage unwanted behavior, but it is not clear how large and widespread the problem is.
