Reliability
This is a gauge of domestic evenness, which comes in handy when quantifying something at a consistent rate. It should be noted that this procedure is comparable to precision. It is enforced in several ways, some of which are discussed below.
Test-Retest
Here, a researcher issues the same examination to a similar set of respondents on separate occasions twice. Afterward, the results are correlated in a bid, to determine the coefficient of stability. The study of these results will determine the stability of these tests over a given time frame. The main disadvantage of this method is the fact that the result varies concerning time; hence similar parameters may produce different results if the timescale is altered. The principal benefit of applying this technique is saved costs since the same test is applied to an existing group of respondents (Trochim, 2006).
Split-half reliability
Here, a test is issued to respondents who are divided into two halves. The scores are categorized separately, and the score of one half is compared to that of the other. Lastly, a comparison of the results is studied to determine the reliability of the study. It is advantageous since it has lower cost and time implications. The main challenge posed is the lack of a proper way by which the questions can be separated before they are administered (Trochim, 2006).
Parallel and alternate forms
Parallel forms
When dealing with parallel tests of reliability, two examinations covering the same area of study are formulated in a similar format. The questions should be of comparable difficulty since they will be distributed at random whilst preparing the tests. Both question papers are then administered to a set of respondents, whose feedback is analyzed for conclusions. A break is allocated between the tests, to prevent instances of fatigue by the respondents (Trochim, 2006).
Among the notable weaknesses of this method, is the fact that the response received is subject to motivation and fatigue levels of the examinees. In addition, the cost implications for creating similar tests are time-consuming and expensive. On the contrary, this method is advantageous, since tendencies of memorization are lowered substantially. The method is also flexible, hence can be used together with other tests (Trochim, 2006).
Alternate forms
This method bears semblance to the test-retest method, save for the fact that a different form is administered for the second time in this case. The examinations are set to ensure close resemblance in the content. After they have been administered, the scores are correlated to produce a coefficient of equivalence. High values imply reliability since a majority of the items will have assessed similar traits. It may be disadvantageous if the time between the tests is lengthy, due to cost implications. On the other hand, the test has greater integrity since the outcome of the subsequent test is not easily alterable.
Validity
It deals with assessment methods employed in the process of psychosomatic and instructive testing and the magnitude to which they are effective. It deals with the magnitude by which substantiation and conjecture react to feedback derived after the use of tests. This makes it comparable to the accuracy.
Face validity
It is applied to determine the relevance of a test concerning satisfying selected criteria. A decision is arrived at based on the face value represented by the test. It is advantageous since it is flexible; hence, offers the best commencement spot for further analysis. On the flip side, it is unreliable, with experts advising disregard for this test when carrying out analysis.
Content validity
This applies to those articles on the test which represent an entire range of possibilities the test should incorporate. This implies that a single test question may be sourced from a single pool of items covering a broad area of topics. The process is time-consuming since it is entirely dependent on joint opinions and this may take time before consensus is achieved. On the contrary, the main advantage of this method is certainty about the questions settled upon, since the deliberations are carried out independently.
Criterion-related
A test exhibits this trait when it exhibits efficacy in determining pointers to a construct. It often assumes two forms, with concurrent validity taking place when criterion measures are realized simultaneously with the outcome of the exam. Predictive validity occurs in the event returns are extracted sometime after the completion of the test. This test is advantageous since it is easily applicable in control circumstances. the main shortcoming is in determining the suitable test for a given scenario, given the two examples mentioned above.
Construct
This is depicted by tests whose results bear some semblance with the forecast and speculative character. Intelligence tests depict some of the daily application methods that require construct validity. The main benefit derived from using this test is the fact that irregularities are easily identified, hence diminished error margins. It; however, cannot be used in instances that lack a projected set of expected results.
Importances of focusing on construct validity and reliability
I believe that adequate focus should be placed on construct validity and reliability. This is because the other methods are deficient concerning determining the legitimacy of experimental procedures of the applied notional concepts that relate to social sciences. In addition, construct validity and reliability tests are quantitative; hence, will encompass all constituents of the construct. Lastly, the test is easy to monitor, since both the construct and the indicator are measurable (Carmines & Zeller, 1986).
References
Carmines, E. & Zeller, R. (1986). Reliability and Validity assessment. California, CA: SAGE Publishers.
Trochim, W. (2006). Types of reliability. Research Methods.