VALIDITY AND RELIABILITY

VALIDITY AND RELIABILITY

1. VALIDITY

   • In the evaluation, it states whether the measurements obtained are valid.
   • Content-Related Evidence
     • The validity of a given assessment is made by comparing the performance of two groups of students.
     • Requirements
       • Sampling of the subject matter from which conclusions will be drawn.
       • The appraisee's performance conforms to the behaviors to be measured.
   • Objetive
     • Evaluate the knowledge imparted in a course in the most appropriate way.
   • Criterion-Related Evidence
     • To compare test results, there must be an overall indicator, either teacher judgments or results.
     • Criteria-measure within the evaluation
       • The assessment has criterion validity as long as its results correspond to those of the criterion measure.
   • Contruct-Related Evidence
     • It is the underlying knowledge and skill that the assessment is intended to gauge.
     • The tests related to the construct comprise
       • a wide range of both knowledge and skill in a specific area
   • Consequential Validity
     • These are the consequences derived from primary factors of the assessment:
       • Accuracy of the measurement of the criteria.
       • Impact on examinee preparation.

2. RELIABILITY

   • It is the consistency of the information obtained from a test.
   • Student Related Reliability
     • These are issues that can make an evaluation unreliable:
       • Fatigue
       • Anxiety
       • Physical factors
       • Psychological factors
   • Rater Reliability
     • It is the differential degree
       • to which several evaluators or the same evaluator at two points in time have obtained consistent scores.
     • Inter-rater reliability;
       • When two or more raters have evaluated the same data.
     • Intra-rater reliability:
       • A single evaluator is in charge of scoring the same data.
   • Test Administration Reliability
     • Classroom conditions could be disturbing the correctness of a test performance.
   • Test Reliability
     • Sometimes the design of a test tends to cause measurement errors.
     • The length of the test, which can leave the test taker feeling unenthusiastic.
     • Two correct answers could be given.