Academic journal article Modern Journal of Language Teaching Methods

The Evaluation of Construct Validity of Iranian Final National Test of English at High School

Academic journal article Modern Journal of Language Teaching Methods

The Evaluation of Construct Validity of Iranian Final National Test of English at High School

Article excerpt

Introduction

Construct validity is required to use performance to infer the possession of certain psychological traits or qualities. These are all hypothetical qualities, called constructs, which are assumed to exist in order to account for behavior in many different specific situations. To describe a person as being highly intelligent, for example, is useful because the term suggests it a series of associated meaning which indicate what his behavior is likely to be under various conditions (Birjandi, 2010).

Before interpreting test scores in terms of these broad behavioral descriptions, however, language teachers and test developers must first establish the constructs which are presumed to be reflected in the test scores actually do account for differences in test performance. This process is called construct validation . In determining construct validity, the aim is to identify all factors which influence test performance and to determine the degree of influence of each. The process includes the following steps: 1) identifying the construct which might possibly account for test performance, 2) formulating testable hypotheses from the theory surrounding each construct, and 3) gathering data to test these hypotheses (Brown, 1997).

The English language is a required subject which is taught in Iranian high schools and as a result, English text books are naturally developed by Iranian text book developers in which reading skills and grammar are emphasized. The students are evaluated on these text books by teacher made tests. These students must take teacher- made tests twice before sitting for the national final test (Website of MoE. Iran).

The national final exams are administered centrally. To come up with a standardized test, a few competent and qualified teachers are invited to design the test in the Ministry of Education. In addition, there is an office inside the Ministry of Education responsible to plan, design, copy, and distribute the papers ( Website of MoE, Iran).

However, the problems which are frequently reported in relation to the final national test of English for third graders in Iran are (1) inadequacy of number of item numbers, (2) poor item wording, and most importantly lack of (3) piloting, item analysis, reliability, and validity studies. These are the concerns that are commonly addressed in the validation process. In order to make a comparison between two national final exams belonging to different school years, four characteristics of these tests i.e. item facility, item difficulty, item discrimination and construct validity were studied (Brown, 1997).

As one way of assessing the construct validity of a test is to correlate its different test components (Alderson & Clapham, 2000), the present study attempted to focus on correlation item analysis of the 2000 and 2014 versions of final national grade 3 in Iranian senior high schools in order to see how it has changed over time. The study, therefore, addresses the following research questions:

RQ 1: How do the test item types, descriptive statistics, item facilities, item discriminations of the 2000 and 2014 versions of Iranian final national tests differ?

RQ 2: How well do the 2000 and 2014 versions of Iranian final national tests correlate?

RQ 3: How well does the total score of the 2014 versions of Iranian final national tests correlate with every subtest?

RQ 4: How well do the different subtests of the 2000 version of Iranian final national test correlate with each other?

Review of the Related Literature

Construct validation is the process of gathering evidence to support the contention that a given test indeed measure the psychological construct the markers intend it to measure. The goal is to determine the measuring of scores from the test, to assure that the scores mean what we expect them to mean (Bachman, 1990). According to Ebel and Frisble (1991), construct validation is the process of gathering evidence to support the contention that a given test indeed measures the psychological construct the makers intend it to measure. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.