Assessment and testing

Validity of tests

A test is valid if it actually measures what it was designed to measure (i.e. the objectives that were agreed on or what had been taught)

Content validity

A test is valid if it adequately and sufficiently measures the particular kowledge, skills or behaviour it sets out to measure. (tennis and running)

Achievement tests

A test designed to measure how much of a language learners have successfully learned with specific reference to a particular course, textbook, or programme of instruction, thus a type of criterion-referenced test. It is typically given at the end of a course, whereas when administered periodically throughout a course of instruction to measure language learning up to that point, it is alternatively called a progress test.

Authenticity of tests

A test is authentic if the language in the test is as natural as possible, the items are contextualized rather than isolated, the topics are relevant and meaningful for thelearners and the tasks represent real-world tasks.

Aptitude tests

A test usually given before actual learning to predict a person's future success in learning a foreign language. It measures sound coding ability, grammatical coding ability, inductive learning ability, memorization.

Student-related reliability

A test is reliable if physical or psychological factors which make an "actual" score deviate from a "true" score e.g. temporary illness, fatigue, a bad day, anxiety etc. are not present.

Test reliability (internal consistency reliability)

A test is reliable if the items or parts of a test are homogeneous, equivalent or consistent with each other.

Face validity

A test is valid if it appears to measure the knowledge or abilities it claims to measure, based on the subjective judgement of an observer or a test taker. e.g. it has familiar tasks, it is doable within the time limit, it has content validity, it presents a reasonable challenge.

Test

Any prepared administrative procedure for measuring a student's ability, knowledge or performance that takes place at identifiable times under time constraints.

Five types of validity

Content validity, criterion-related validity, face validity, construct validity, consequential validity

Admission tests

High-stakes tests taken to provide information about whether a candidate is likely to succeed

Proficiency tests

Regardless of previous training they measure general ability in a language.

Formal assessment

Systematic tests, tasks or procedures given under conditions that ensure the assessment of individual performance in any given area.

Practicality of tests

An effective test is practical, inexpensive, easy to administer, takes time constraints into consideration and has a scoring procedure that is specific and time-efficient.

Progress tests

Tests prepared by a teacher and given at the end of a chapter, course, or term are progress tests. Progress tests may be regarded as similar to achievement tests but narrower and much more specific in scope. They help the teacher to judge the degree of success of his or her teaching and to identify the weakness of the learners.

Informal assessment

Unplanned comments, verbal feedback on a student's performance under normal classroom conditions rather than through the use of standardized tests.

Five Principles of Language Assessment

positive washbach effect, practicality, reliability, validity, authenticity

Seven types of tests

proficiency, diagnostic, placement, aptitude, admission, progress, achievement

Washback effect of tests

A beneficial or harmful effect of tests and exams on teaching and learning.

Summative test

A high stakes test given at the end of a course of instruction with the aim of evaluation.

Diagnostic tests

A low-stakes test that is designed to provide information about L2 learners' strengths and weaknesses to benefit future instruction.

Formative test

A low-stakes test that is given during the course with the aim of monitoring the learning process and identifying students' strengths and weaknesses.

Reliabiliy of tests

A reliable test is consistent and dependable. The same test given to the same students in two different occations should have the same results.

Norm-referenced test

A standardized test (e.g. TOEFL) that measures how the performance of a particular test taker compares with the performance of other test takers so they are placed on a mathematical continuum in rank order.

Assessment

A systematic approach to collecting information and making inferences about the ability of a student or the success of a course. Assessment can be done by tests, interviews, observations, etc. i.e. any kind of evaluation is assessment.

Test administration reliability

A test is reliable if good conditions in which the test is administered are provided. e.g. no background noise, new/high quality equipments, chairs/tables, quality of photocopies etc.

Criterion-related validity

A test is valid if it is correlated with an established external criterion measure. (criterion measure: a standard or objective that has been agreed on)

Criterion-referenced test

A test that gives test takers feedback on specific objectives (criteria). The test taker must reach a previously agreed level of performance to pass the test.

Placement tests (also: entry tests)

Tests given to assign students to classes/programmes appropriate to their level of proficiency.

Scorer reliability (inter-rater reliability)

A test is reliable if different examiners or judges making different subjective ratings of ability (e.g. of L2 writing proficiency) agree in their evaluations of that ability to avoid human error, subjectivity and bias.

Four types of reliability

Student-related reliability, scorer reliability, test administration reliability, test reliability