PART THREE
CROSS-CASE ANALYSIS AND DISCUSSION
8.2 Cross-case Analysis
8.3.5 Enhancement on the MOBIlearn task model
73
Fall, 2001 administration of the SAT, 60% or roughly 159,940 students first eliminated incorrect alternative before selecting one alternative as the answer. Likewise, Dirkwager (1996) also discovered that most students used this strategy. Thissen (1976) noted that students who do not possess the necessary knowledge to answer an item do not randomly select an answer, but rather use partial knowledge to first eliminate the incorrect and select one of the remaining option. Tomkowiez(2000) also shares his view.
According to Mehrens and Lehmann (1991), logic and evidence suggest that it is informed guessing that predominates. Dowing(2003) further argued that informed guessing should not be discouraged because it is a reality of life. Much of what individual does throughout is in fact based on incomplete knowledge. Although frequently used by students, the strategy of first eliminating incorrect option before selecting the answer is only regarded as a test taking strategy to maximize test score.
It has not been formalized as a scoring method. According to Item Response Theory (IRT), the test information is simply the sum of each item‟s information. This means that we can create a test to have any kind of information we like by judiciously choosing our items.
74
of students‟ attributes such as achievement, aptitude, interest, values, personality and intelligence. These attributes exist in the abstract as psychological constructs. Their presence can only be inferred indirectly through their external manifestations. The reliability and accuracy of a test depends on how accurately you can describe the attribute in the question. The description is done through test items because each item calls for a specific behavior. The test items depict the attributes to be measured but test scores have no proportional relationship to the attributes measured. This is because a test does not measure directly. Okoli (2000) concluded that there is no guarantee that the test items actually describe the attributes.
Ebel and Damrin (1990) noted that essentially, there are four bases on which candidates respond to multiple choice questions and they are: direct knowledge, testwiseness, response set and chance guessing.
Direct Knowledge: In this case, an individual‟s response is based on his information about the content of the question and as such the decision to select an answer is always knowledge- based. The answer may be selected with complete knowledge and unerring certainty or only enough knowledge to arrive at the answer through vague, almost intuitive reasoning.
Testwiseness: In this case, an individual‟s response is based upon his knowledge of the design and format of the objective tests.
75
Response Sets: In this case, the subject‟s response is based on an ingrained set of personality habits which are independent both of his knowledge of the content of the question and of his testwiseness.
Chance Guessing: This refers to choosing an answer exclusively by chance and is also called blind guessing. This strategy allows the examinee to gain points beyond that which he/she would have received on the basis of sure knowledge of the subject matter.
Millman, Bishop andEbel (1985) made a distinction between informed and blind guessing. Informed guessing is one where the examinee can eliminate one or more options as incorrect and then choose from the remaining options. Ability to eliminate some options increases the probability of selecting the right option. It must be noted that the ability to eliminate some options implies partial knowledge of the subject matter. Blind guessing on the other hand is selecting an answer at random without considering the content of the options.
Davis (1984) opined that testees select their answers on one or more of the following bases:
(i) Sufficient knowledge to identify the correct choice
(ii) Partial knowledge that permits choice followed by guessing among all of the remaining choices or by resorting to irrelevant considerations like characteristics irrelevant to content of the item.
(iii) Guessing among all the choices after considering the item as a whole.
76
(iv) Partial misinformation that leads to elimination of one or more choices, including the correct choice, followed by guessing among all of the remaining choices or by resort to irrelevant considerations like those listed in the next point below
(v) Characteristics irrelevant to content of the item, such as length of the choices, precision of wording of the choices, number of times each position (A, B, C, D.E, etc.) has been marked as a correct option, pattern of choice positions that appears on the answer sheet.
(vi) Sufficient misinformation to identify as corrects, one of the incorrect choices.
(vii) Random marking of choices in items that have not even been read. This refers to blind or chance guessing which normally occurs when a testee is under examination tension with little time left to leave the examination hall or when he/she has little or no idea of the subject matter.
Nenty (1987) remarked that when an examinee encounters a test item, the intention is that he will respond to the item within the best of his ability, that is, on the basis of how much of what the item measures that he possesses. In an ideal situation, the factors in the encounter are: the ability of the examinee in what is being measured and the difficulty of the item.
According to Nenty (1987), an examinee enters a testing situation with a variety of background characteristics besides his level of ability on what the test is trying to measure. Such characteristics include: his level of anxiety, carelessness, motivation and mood, his level of testwiseness and ability to guess, his tendency to copy from
77
others, his level of exposure to and familiarity with related subject matter, his disposition to the type of item format used in the test and the individual‟s idiosyncratic cognitive mode. Similarly, items on the test enter the test situation with a variety of characteristics other than its ability to elicit that behavior the test was designed to measure. These include its level of demand on skills and knowledge other than that which the test was designed to measure, its level of bias, the order of presentation of items, the specificity of the test and item directions and its level of ambiguity.
A closer review of the factors highlighted by researchers (Ebel andDamrin, 1990;
Davis, 1984; Nenty, 1987) shows that apart from true/direct knowledge other factors such as the nature of the test item, chance guessing, testwiseness and anxiety have significant roles to play in students‟ academic achievement.
It is clearly evident that guessing strategy and testwiseness deserve much more attention than they are presently being given in schools because it is an important construct which affects test scores hence training in it may be used to improve test scores of Students in Mathematics.