Data mean the actual scores or factual information that are obtained from the examinees on administering a test or questionnaire. Only on the basis of these data conclusions are drawn. The obtained data that are not undergone logical or statistical analysis, are often referred as 'raw data'. The data so obtained from non-experimental observation are known as 'observational data'. The format which issued to record the obtained data in a systematic way is called 'data sheet'.
Debating is considered as a tool for assessing the presence of mind and art of oration of an examinee. It is an argument type of speech delivered in a formal manner on the positive and negative aspects of a topic, before the audience. The formal debating assumes new patterns of educational value requires to be developed and thereby evaluated in the formal education system.
Decile is a type of percentile rank (norm) like quartile, but with less common usage. It divides the distribution into ten equal parts (bands) with each part containing 10 percent of the norm group. Sattler [1982, p.634] clarifies it as "a score in a frequency distribution below which either 10% (1stdecile), 20% (2nd decile), … 90% (9th decile) or 100% (10th decile) of the total number of scores fall". [Also, see, Quartile].
Decile Point Test is one of the computer adaptive tests. This kind of tests are used in a situation where decision of admission or exemption is involved. Depending on the decision the computer presents the candidates only such items the difficulty level of which is predetermined as per the actual requirement. It is considered highly efficient and accurate for decision making purposes. This test is useful to [Henning] maximise decision accuracy by loading it with items whose information index is highest at the decision points along the ability continuum.
Degrees of Freedom are used to denote the number of independent comparisons which can be made between the members of a sample. They are [Henning] the number of units of analysis after loss due to statistical computation. With correlation, two degrees of freedom are lost to position the regression line. With chi-square, d.f. loss equals rows plus columns minus one.
Demographic Questionnaire is one of the tools used for eliciting information of personal nature, such as age, sex, place of birth, languages known, educational background, and so on from the respondents. Kinds of information elicited depend on particular research programme needs. The use of this questionnaire is quite common in behavioural sciences. Unlike bio-data, (which is always presented by the candidates themselves), demographic questionnaire may sometimes be filled in by the investigators also as they have got a wider coverage.
Derived Scores are the scores that are obtained from the raw scores by an arithmetic operation. They vary from one another in their usefulness. If different sets of marks are to be added or to be compared, they need to be scaled to a common standard where such standard is lacking. For converting raw scores into derived scores, examination scores can be expressed by [Natarajan, 1977c, p.52]:
(1) Comparison with an 'absolute standard',
(2) Inter-individual comparison, or
(3) Intra-individual comparison.
Item (1) can be :
(a) Percentage correct scores.
(b) Letter grades.
Item (2) can be :
(a) Based on mean standard deviation -
Z = Scores, T = Scores, AGCT Scores, CEEB Scores
(b) Based on rank within groups -
Ranks and bands, Percentile ranks, Letter grades,
Normalised standard scores, viz., T-scaled score,
Stanine score, C-scaled score.
[c] Based on range of scores in a group - Percent placement.
Item (3) can be :
(a) Ratio IQS
(b) Intellectual status index.All these types represent different methods of expressing one's performance. Since most of our examinations involve inter-individual comparison, we ought to know various methods of converting raw scores to derived scores to make them meaningful. The Z score is basic to all other derived scores based on mean and standard deviation.
Deviation refers to the variability of scores. It indicates the spread of the scores around the mean. The greater the value of deviation, the wider the spread of scores around in the mean is. In other words, subtract the mean of the scores from each score in turn. The differences obtained by subtraction are called deviation.
Deviation IQ is the Intelligent Quotient (IQ) obtained by converting raw scores on an intelligence test to a score distribution having a mean of 100 and standard deviation of 16 (in case of Stanford-Binet) or 15 (in case of Wechsler Tests).
Diagnosis is one of the most important uses of evaluation. The very purpose of testing is to diagnose (or to locate, or to identify) the strengths and weaknesses of learning in order to suggest remedial measures for necessary alteration in the process of teaching/learning. The other purposes are monitoring, guidance, prediction, selection, and grading which are separately explained.
Diagnostic Evaluation is concerned with the identification of the learning difficulties of recurring nature or areas of weaknesses that are left unresolved by the standard corrective procedure of informative evaluation. The primary aim of this evaluation is to determine the cause of such learning problems or weaknesses and to formulate a plan for remedial measures.
Diagnostic Test is designed to measure the strengths and weaknesses of an examinee in a particular area of learning, such as reading or arithmetic. It provides critical information not only to the students, but also to the teachers and educational administrators for taking suitable remedial measures. Sattler (1982), however, views it as an achievement test composed of items in a number of sub-areas, with the purpose of diagnosing on individual's relative strengths and weaknesses in the sub-areas.
Diary is a daily record of events. This method of observation attempts to record all happenings in a given period of time with reference to the object of observation. Although similar to anecdotal method, diary record is certainly a different one in view of its attempt to record every thing that is considered important. Diary record is mostly maintained by the individuals themselves for their references/ benefits. However, such records are also maintained by the monitors of programmes to know the daily performance of the clients.
Dichotomy refers to a division of two groups or two constituents of a population, or sample which are usually the measurable variables.Dictation Test is the one which requires an examinee transcribe a passage he hears read (live or on tape). In language teaching/learning process, it has been in practice, irrespective of the situation whether L1 or L2 or L3. It measures various aspects of the progress and problems in language learning besides the listening comprehension.
Difficulty Level (Difficulty Index) is one of the characteristics of a test item obtained on the basis of item analysis. Item difficulty, difficulty level, difficulty index, facility value are all different terms that are used synonymously to represent the same characteristic. The difficulty level is communicated as the proportion of the number of persons who answer a particular item correctly with those who merely attempt it. Thorndike and Hagen (1977, p.59) view it as "no definition is absolute, all ultimately are relative to the performance of test takers. An item that is quite advanced and complex for fourth graders would probably seem quite simple to sixth graders". However, with a little variation Ebel (1979) clarifies it as the proportion of examinees in a group it as the proportion of examinees in a group who do not answer the test item correctly. The most common index of difficulty is the percent of incorrect response. But Sattler (1982) considers it as the proportion (percentage of individuals passing a given test item; the larger the index, the easier the item.
Difficulty Scale is a test or scale (English, Horce B. and Ava C. English) whose scoring units are tasks graded in order of difficulty. The individual may be given the rank score of the most difficult item passed, or of the least difficult item failed.
Difficulty Scale is the percentage of some specified group such as the examinees of a particular age or class who answer a test item correctly.
Definition Validity refers to the postulated validity of a test when the items that constitute test themselves are taken as the definition of the variable to be measured. If a test contains items on cognate vocabulary, non-cognate vocabulary, technical vocabulary, it can arbitrarily be called a test of vocabulary. But it is seldom used/in practice. Definitional validity is also a measure of content validity.
Dimension refers to any measurable extent or magnitude or characteristic by which object or event can be positioned in a quantitative series.
Directed Speaking Tests measure the ease and accuracy with which the student handles the patterns of the language being learned. These tests are quite common in all the language teaching situations.
Direct Measurement is one in which the observer perceives on a scale the quantitative statement of the value. Grocery measured in a balance, time seen in a clock, etc., would be appropriate examples for Direct Measurement. [Ant.: Indirect Measurement].
Direct Test/Questioning is designed to measure a particular (language) behaviour in a natural way and in the real context. For example, conversing with a person or interviewing him may be considered as more direct than a dictation test for measuring the language proficiency.
Discrete Grammar Items are the items that are used for the evaluation of the passive knowledge of grammar only. These items are usually in the multiple choice question format.
Discrete Point Test is the one that attempts to measure only one point at a time usually in a multiple choice item format. Although a similar view is expressed by Oller, Jr. and Perkins (1978, p.39), they have their own reservations about this test. As per them "when discrete point test requires a person to make a phonemic contrast, inflect a verb, or define a lexical items in isolation and without any (of these stated above) organizational constraints. 'Certain crucial properties of language are necessarily lost". Therefore, in place of discrete point tests, a 'special class of integrative tests, namely, 'pragmatic tests' have been suggested by Oller [see, Pragmatic Test]. Carroll and Hall [1985, p.166] clarify it in detail as "tests consisting of items which aim to deal with one point at a time, e.g., the 'skill' (e.g., listening) and one language phenomenon (e.g., the distinction between the minimal pair 'sit and bit'). The items tested will have no meaning relationship with each other and will not be related to any overall communicative purpose". In other words, this test employs items measuring performance over a unitary set of linguistic features. An example would be a multiple choice test of article usage.
Discriminability is the capacity of an item to discriminate between the higher ability students and the lower ability students. The value of this capacity is known as the discrimination power of that particular item.
Discrimination Index (DI) indicates the ability of a test item to discriminate between the top scoring and bottom scoring students. Discrimination is usually measured by the correlation between the score on the item and the score on the total test. The higher the correlation value is the better the discrimination power of the item. Another method of computing item discrimination power is to compare the responses of the top item. The procedure is as follows: On the basis of the total score, two groups of students are selected (with equal size of top and bottom scoring groups). For each item the proportion of bottom scoring students who get the item right is subtracted from the proportion of top scoring students who get the item right. The result arrived at is Discrimination Index or Index of Discrimination. For example, if 0.75 of the upper group get an item right and 0.25 of the lower group get it right, the item has a DI of 0.5. the higher the value of index is the greater the discrimination index ranges from +1.00 to -1.00. Items with positive discrimination indices discriminate in the same manner as the total score does. Items having the DI value of +0.30 and above, are considered as having the satisfactory degree of discrimination power, and hence they could be retained. The items with negative value have to be discarded.
Discriminant Validity is a kind of construct validity established with reference to a multitrait-multimethod validation matrix.
Discussion is also an examination and comparison of the views of two or more persons each of whom contributes to general understanding. Discussion is mingled in varying degrees [English, Horce B. and Ava C. English] with argument, persuation, exhibitionism and striving for dominance, all of which are subversive of discussion. One's growth in the scholastic as well as non-scholastic areas can be judged/assessed through this method of examination.
Dispersion is the degree of scatter or clustering shown by observations. It is usually measured as an average deviation about some central value like mean deviation, standard deviation, etc.
Distractors are the incorrect answer options. They are 'attractors' to the lower ability (poorly informed) students. Mostly they find place in multiple choice items.
Distractors Tally is a tally of the response frequencies for each of the multiple choice distractors. It is used to improve the quality of items by ascertaining the effectivity of the distractors. If a distractor does not attract any response, it means that it is not at all functioning as an effective distractor and hence it should be replaced with a more attractive option. In order to locate this kind of weaknesses of items, distractor tally helps a lot.
Distribution (of Scores) is a set of tabulated scores or marks, wherein the frequency of occurrence of each score is shown.
Domain Referenced Test [see, Criterion Referenced Test].