Number Of Test Items Research Articles

Intra-Oral Recognition of Geometric Forms by Normal Subjects

Oral form recognition abilities were determined for 40 normal Ss between the ages of 20 and 59 yr., using 12 different geometric forms in 8 sizes. The purposes were: (1) to explore such related variables as form complexity, form size, Ss' age, sex and education level and Ss' response time; and (2) to reduce the total number of test items to a more manageable level by selecting those shapes and sizes from the original test items which contribute most to the obtained oral stereognosis scores. There were no significant differences for sex or educational level, but significant relations were found among performance levels and age groups. In addition, Ss' performance tended to be inversely related to response time. Within the limits of several criteria, 10 forms were selected which can practically and effectively provide a measure of oral form recognition. And, these forms may permit assessment of oral sensory integrity.

On some sequential life tests

B. Epstein and M. Sobel proposed sequential life tests in the exponential case [1], in which a statistic V(t), called "total life", is observed continuously in time t and decision is made at an instant when V(t) does cross the preassigned limits. In the nonreplacement case of tha t test, the probability that a decision can be made af ter all ~ items on test have failed is smaller than one. Therefore the average test time can not remain to be finite so long as ~, the number of test items simultaneously placed on test, is finite. Thus they recommended the sequential test, in which n must be determined so large tha t the probability not to reach any decision is negligible small and some decision rules must be defined in advance to provide for such indeterminable cases. However, it seems to be expensive and troublesome that a sufficiently large number of items are placed on test simultaneously, because test equipments must be so large according to the number of test items. In this paper, the test procedures applicable to the wider class of distributions including the exponential case will be proposed, in which a suitable number of test items are placed on test repeatedly untiU it reaches a decision. The operating characteristic function, the average failure (sample) number and the average test time will be obtained in comparison with the usual sequential test without a continuous time parameter.

Sampling Error due to Choice of Split in Split-Half Reliability Coefficients

Sampling Error due to Choice of Split in Split-Half Reliability Coefficients

REDUCING TYPE‐2 SAMPLING ERRORS IN THE ESTIMATION OF TEST RELIABILITY

ABSTRACTThe formula and derivation are given for the sampling variance of the “random‐halves” reliability coefficient when items, rather than examinees, are sampled. When the number of test items is large, this sampling variance is found to be larger than the Kuder‐Richardson formula‐20 coefficient by a factor equal to the number of test items. The usual matched‐halves reliability coefficient will likewise usually have a large sampling variance, since it likewise is based on an arbitrary split of the test items. An adaptation of the Jackson‐Ferguson “battery reliability coefficient,” used as an estimate of the matched‐forms reliability of the test, avoids this relatively large sampling variance.

Approximating maximum test validity by a non-parametric method

The Gleser-DuBois conditions for selecting from a number of test items those which will maximize the correlation between total test score and criterion will degenerate into expressions requiring only item counts on total distributions and the upper halves of distributions. A grouping convention for scores near medians is recommended. The inefficiency of the method is easily compensated for, because, regardless of the size of the sample, only standard test-scoring equipment and brief computations are required. A procedure is outlined, and some applications are discussed.