12.01.2015 Views

RESEARCH METHOD COHEN ok

RESEARCH METHOD COHEN ok

RESEARCH METHOD COHEN ok

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

COMMERCIALLY PRODUCED TESTS AND <strong>RESEARCH</strong>ER-PRODUCED TESTS 417<br />

they may not be exactly suited to the purpose<br />

required. Further, several commercially produced<br />

tests have restricted release or availability, hence<br />

the researcher might have to register with a<br />

particular association or be given clearance to<br />

use the test or to have copies of it. For example,<br />

Harcourt Assessment and McGraw-Hill publishers<br />

not only hold the rights to a world-wide battery<br />

of tests of all kinds but also require registration<br />

before releasing tests. In this example Harcourt<br />

Assessment also has different levels of clearance,<br />

so that certain parties or researchers may not be<br />

eligible to have a test released to them because<br />

they do not fulfil particular criteria for eligibility.<br />

Published tests by definition are not tailored<br />

to institutional or local contexts or needs; indeed<br />

their claim to objectivity is made on the grounds<br />

that they are deliberately supra-institutional. The<br />

researcher wishing to use published tests must be<br />

certain that the purposes, objectives and content of<br />

the published tests match the purposes, objectives<br />

and content of the evaluation. For example,<br />

a published diagnostic test might not fit the<br />

needs of the evaluation to have an achievement<br />

test; a test of achievement might not have the<br />

predictive quality that the researcher seeks in an<br />

aptitude test, a published reading test might not<br />

address the areas of reading that the researcher<br />

is wishing to cover, a verbal reading test written<br />

in English might contain language that is difficult<br />

for a student whose first language is not English.<br />

These are important considerations. A much-cited<br />

text on evaluating the utility for researchers of<br />

commercially available tests is produced by the<br />

American Psychological Association (1999) in the<br />

Standards for Educational and Psychological Testing<br />

(http://www.apa.org/science/standards.html).<br />

The golden rule for deciding to use a published<br />

test is that it must demonstrate fitness for purpose.If<br />

it fails to demonstrate this, then tests will have to<br />

be devised by the researcher. The attraction of this<br />

latter point is that such a ‘home-grown’ test will<br />

be tailored to the local and institutional context<br />

very tightly, i.e. that the purposes, objectives<br />

and content of the test will be deliberately<br />

fitted to the specific needs of the researcher in<br />

aspecific,givencontext.Indiscussingfitnessfor<br />

purpose, Cronbach (1949) and Gronlund and Linn<br />

(1990) set out a range of criteria against which a<br />

commercially produced test can be evaluated for<br />

its suitability for specific research purposes.<br />

Against these advantages of course there are several<br />

important considerations in devising a ‘homegrown’<br />

test. Not only might it be time-consuming<br />

to devise, pilot, refine and then administer the<br />

test but also, because much of it will probably be<br />

non-parametric, there will be a more limited range<br />

of statistics that may be applied to the data than<br />

in the case of parametric tests.<br />

The scope of tests and testing is far-reaching;<br />

no areas of educational activity are untouched<br />

by them. Achievement tests, largelysummativein<br />

nature, measure achieved performance in a given<br />

content area. Aptitude tests are intended to predict<br />

capability, achievement potential, learning potential<br />

and future achievements. However, the assumption<br />

that these two constructs – achievement<br />

and aptitude – are separate has to be questioned<br />

(Cunningham 1998); indeed, it is often the<br />

case that a test of aptitude for, say, geography at a<br />

particular age or stage will be measured by using an<br />

achievement test at that age or stage. Cunningham<br />

(1998) has suggested that an achievement test<br />

might include more straightforward measures of<br />

basic skills, whereas aptitude tests might put these<br />

in combination, for example combining reasoning<br />

(often abstract) and particular knowledge; thus<br />

achievement and aptitude tests differ according to<br />

what they are testing.<br />

Not only do the tests differ according to what<br />

they measure, but also, since both can be used<br />

predictively, they differ according to what they<br />

might be able to predict. For example, because an<br />

achievement test is more specific and often tied to a<br />

specific content area, it will be useful as a predictor<br />

of future performance in that content area but will<br />

be largely unable to predict future performance<br />

out of that content area. An aptitude test tends<br />

to test more generalized abilities (e.g. aspects of<br />

‘intelligence’, skills and abilities that are common<br />

to several areas of knowledge or curricula), hence<br />

it is able to be used as a more generalized predictor<br />

of achievement. Achievement tests, Gronlund<br />

(1985) suggests, are more linked to school<br />

Chapter 19

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!