Online assessments are a mirror image of their offline and traditional counterparts. Traditional tests, also known as paper-pencil tests, are administered on individuals to assess their abilities, intelligence, to unravel personality traits, to measure their aptitude, etc. Similarly, online tests can measure any of the qualities mentioned above with online administration as the chief differentiator. The only differences between traditional testing methods and online tests are the use of Internet as a platform, and of course the various benefits that come with online testing.
What is a test actually?
A test can be defined as a series of questions on basis of which some sort of insight is sought. However, in psychology or in education, test is something a lot more than this basic definition. A.K. Singh defines testing as a “standardized procedure to quantitatively or qualitatively measure one or more aspects of a trait by means of a sample of verbal or non-verbal behavior”.
Characteristics of a good test
If you are looking for a custom online test to hire employees, there are many characteristics you might want to keep in mind. These characteristics are similar to those of traditional tests and in this post, we shall take a look at all of these characteristics. Ensuring that an online test has all these qualities will help you to assess employees in a scientific and statistically recognized manner. Test construction is a science that borrows from social sciences, statistics and probability. With this in mind, it helps to know the characteristics of good online tests, which we have listed below.
Traditionally, five characteristics are listed in statistics textbooks and test construction manuals. Those five characteristics are explained below:
1) Test Objectivity
In this case, objectivity refers to the mutual agreement between experts that the test is objective and that it is free from ambiguity. This also implies that experts agree on items and their scoring, which leads to uniformity and fairness while scoring a test.
There are two main entities of objectivity: objectivity of scoring the test and the objectivity of test items. Objectivity of test items refer to the items being perceived in the same way by everyone who takes the test. This means, the items should be presented in uniform manner during the test to all test-takers, regardless of who is taking it. Objectivity of test scoring requires that the scoring of the test is standard and everyone scores it the same way. This uniformity helps to maintain fairness and accuracy of the test results.
In the case of online tests, test objectivity refers to uniformity in the way items in a questionnaire are presented and also, that the software is programmed to interpret the test results uniformly. This is actually taken care of during coding, and one of the benefits of online assessments is that it is almost always objective, even during projective assessments, though by definition projective tests are not objective.
2) Test Reliability
Reliability is defined as a test’s quality of being self-correlative. It refers to the quality of a test yielding consistent results when the test is administered on the same sample within a given period of time. For example, a person who scores 120 on an IQ test may score 122 on a different day. However, he will not score 98 on a third day. In short, reliability is the quality of a test to be able to produce same kinds of reports no matter how many times an applicant takes it.
Internal consistency of the test refers to the consistency of results in a single administration. Temporal consistency on the other hand refers to consistency of test results upon testing and retesting. Test reliability should therefore possess both internal and temporal consistency. A reliable test is free from internal defects and errors, and thus avoids errors of measurement.
Online testing ensures reliability of tests because during custom construction of assessments, computerized techniques are used to ensure that a test consistently produces similar results when administered on the sample (test-taker). Online tests also ensure that there is both internal and temporal consistency in the results derived, which is very important if we want to make hiring decisions accurately and fairly.
3) Test Validity
An important prerequisite of a good test is that it is valid. Test validity is defined as its ability to measure what it intends to measure, when compared with independent and external factors. Alternatively, we can also define test validity as its quality to correlate positively with certain outside criteria. The criterion should be independent and generally considered as one of the best index to measure the trait or ability that the test is measuring.
For example, an intelligence test should measure intelligence. If one were to launch a new intelligence test, it should compare well with a previously established intelligence test, such as the Stanford-Binet Test. If the new test’s correlation coefficient is close to that of the previous test, we can safely say that it is a valid test, which measures intelligence similar to a previously standardized intelligence test. Usually, test validity depends on test reliability. If a test is not reliable, it will not be valid either.
In online assessments, test validity can be ensured by following standard test writing procedures. Test writing is a scientific procedure which ensures that the items of a test are chosen carefully, and tested on a sample, and later standardized on a sample, which helps to provide norms.
4) Availability of Norms
Norms refer to the average performance of a representative sample. This is usually obtained after standardization. During standardization, various groups that represents the population to be studied (also known as a ‘sample’) are studied. The test is repeatedly administered to check for reliability and validity.
Under these test-retest conditions, important results are calculated and stored. Finally, an average of these various performances of the representative sample is established. These averages are referred to as the norms. In a standardized test, an applicant’s results are compared with the norms (in other words, the average of the sample that represents him or her). If he or she deviates too much from the norm, they represent outliers. If they are very close to the norms, then they probably performed averagely.
An online test has the advantage of being standardized and norms developed very easily, thanks to its ability to be administered online. It is easier to gain access to samples (people who represent the population or individuals to be studied and assessed) online, and thus, establish norms. While hiring, we obviously want to choose those applicants who have scored better than the average norms. Norms help companies to choose candidates in comparison with a larger pool of samples, even if they are administering the test to just one individual. If an individual scores very highly on an assessment test, and has a very high percentile with respect to the norms, further assessments can be avoided and he can be hired, reducing costs and time.
Test practicability is nothing but how practical it is to use it as a test. Practicality comes in terms of the time taken to complete it, its length, scoring and interpretation. It should not take too long. However, it is difficult to define how long is too long.
Many projective tests such as the Rorschach Inkblot Test are not practical because they take a long time to interpret. Psychologists often spend hours trying to code al the responses and then interpreting those codes. Exner’s system of scoring is one of the methods to score a Rorschach response sheet. There are several others but all of these have the problem of practicability. They take far too long and are not suitable for regular employment assessment conditions. Moreover, scoring should be such that basic knowledge should be enough to score a test.
Tests like Rorschach and Thematic Apperception Test require clinical psychologists to be present to interpret the data, which is simply not practical. In other words, a good test is short, easy to score, and easy to administer. Online tests can be customized to be very practical and easy. As scoring is done by the computer, it makes things more practical.
A few other characteristics specifically for the online testing environment
Going further, to adapt tests to an online medium, we have listed five more characteristics of a good online assessment test. These qualities of a test help it to be administered effectively online, without the applicant feeling fatigued or without him/her finding it difficult to respond to questions within a given period of time.
6) Test comprehensiveness
When it comes to Offline questionnaires, we often see items (an item is defined as a single question within a test) that do not accurately seek all the information required. A good test should test all the attributes of a trait that is being measured.
For example, an intelligence test that only tests an individual’s verbal ability is not comprehensive. It needs to measure other forms of intelligence such as performance intelligence, reasoning skills, mathematical ability, etc. A comprehensive test lists all the factors of a trait that is being tested, and seeks to measure all those attributes to comprehensively measure what is being measured. This is what is meant by test comprehensiveness.
Online tests are suited for test comprehensiveness because it is easier to factor in all the attributes of a trait that is being measured, and customize the items in such a manner that the administration is practical too, while remaining comprehensive. Usually, practicability of a test is undermined when we make it comprehensive, especially in paper pencil tests. However, in online tests, both practicability and comprehensiveness of the test can be maintained, in order to produce better results.
7) Time appropriate
Time appropriateness of a test is slightly different from it being practical. Under practicability of a test, we discussed that it has to be short in duration, with respect to administration, scoring and interpreting. However, time appropriateness of a test refers to its ability to allow for recruitment given whatever the circumstances of a company are.
For example, one can always deliver a short test that is practical yet comprehensive, but it may not be time appropriate. If the same test is being administered to thousands of job applicants, the entire testing procedure and selection process should be fast too, so that the right candidate is hired without any delays.
When a test is time appropriate, we can also assume that it is already comprehensive and practical. This is not to say that time appropriate tests are short in nature. On the other hand, they help companies to recruit quickly without testing unnecessary traits. In other words, we test only what is required for that particular job profile, yielding results that which help in making the recruitment decision. Using the internet to administer tests help in remaining time appropriate too.
Most importantly, a test should not be exorbitantly priced. It should be within the reach of a company’s expenditure for recruitment assessments. Most of the times, buying licenses to paper-pencil tests cost a lot. This forbids companies from accessing standardized tests that help in making the right hiring decisions.
Good online tests are not as expensive as paper-pencil tests because, you do not have to purchase license separately for each time you administer it. Usually, the costs are included in custom test construction, which includes professional charges to construct the test plus the software program that administers the tests to applicants.
9) Lack of ambiguity
Online tests are usually taken by applicants in the comfort of their own homes. It is not possible for applicants to ask doubts regarding items or take breaks in between, because the items are displayed on the screen only for a certain period of time, unless specifically programmed. This warrants unambiguous test items so that test takers are not left confused or in doubt, while taking it.
This is one of the disadvantages of online test assessments as examiners cannot be present during the time of assessment, reducing the interaction between examiners and test takers. Thus, test items must be carefully selected so that there is no ambiguity at all. A good online test lacks ambiguity and is crystal clear to understand.
Test standardization is a lengthy procedure in which a test is tested on several samples, and norms are developed. This is a very important part of test construction because if a test does not measure what it claims to measure, and if it does not consistently produce similar results when conducted on a group, and if the results cannot be compared with that of a representative group, it cannot be used effectively. In fact, tests that are not standardized cannot be relied upon. Online tests can be easily standardized as samples can be found easily too. It is quicker and easier to conduct pilot studies based on which standardization can happen and norms can be developed.
Important points to consider while developing an online assessment test
It should be now clear that a good test has several important characteristics. It should be objective, to begin with. It should not have any room for personal bias or subjectivity. Secondly, it should be reliable. The test should yield test results consistently, without many changes, when administered on the same individual.
Once it is reliable, we should ensure that a test is valid. Test validity ensures that a test can be compared with other such tests, and that it correlates positively. Norms ensure that a test result can be compared against the average of a representative sample, while practicability ensures that a test is quick and not lengthy and drawn out.
We must also ensure that online tests meet a further few criteria. These include test comprehensiveness, which means it should test everything related to a certain trait. It should also be time appropriate, economical and must not have ambiguity of any sorts. Time appropriateness ensures that large numbers of applicants can be assessed quickly. Economical tests ensure that a company is not overburdened while developing or purchasing tests.
Most importantly, online tests should lack ambiguity as applicants cannot see the examiner face to face usually, unlike in paper-pencil tests. Last but not the least, standardization of tests should be taken very seriously, as the whole edifice of test construction depends on this.
Talk to us today
At Testofy, we realize the importance of ensuring that an online assessment tool meets all the 10 criteria listed above. It takes a lot of effort to develop these tools and a lot more professional skills too. It is very important for us to ensure that tests are reliable and valid, and that norms are developed effectively. A lot of effort goes into building a custom test and we understand that assessments form the backbone of any recruitment strategy. If you are looking for a good online assessment tool to recruit employees at any level, speak to us today.