12.01.2015 Views

RESEARCH METHOD COHEN ok

RESEARCH METHOD COHEN ok

RESEARCH METHOD COHEN ok

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

CONSTRUCTING A TEST 423<br />

where<br />

A = the number of correct scores from the high<br />

scoring group<br />

B = the number of correct scores from the low<br />

scoring group<br />

N = the total number of students in the two<br />

groups.<br />

Suppose all 10 students from the high scoring<br />

group answered the item correctly and 2 students<br />

from the low scoring group answered the item<br />

correctly. The formula would work out thus:<br />

1<br />

2<br />

8<br />

(10 + 10)<br />

= 0.80 (index of discriminability)<br />

The maximum index of discriminability is 1.00.<br />

Any item whose index of discriminability is less<br />

than 0.67, i.e. is too undiscriminating, should be<br />

reviewed first to find out whether this is due to<br />

ambiguity in the wording or possible clues in the<br />

wording. If this is not the case, then whether the<br />

researcher uses an item with an index lower than<br />

0.67 is a matter of judgement. It would appear,<br />

then, that the item in the example would be<br />

appropriate to use in a test. For a further<br />

discussion of item discriminability see Linn (1993)<br />

and Aiken (2003).<br />

One can use the discriminability index to<br />

examine the effectiveness of distractors. Thisis<br />

based on the premise that an effective distractor<br />

should attract more students from a low scoring<br />

group than from a high scoring group. Consider<br />

the following example, where low and high scoring<br />

groups are identified:<br />

A B C<br />

Top 10 students 10 0 2<br />

Bottom 10 students 8 0 10<br />

In example A, the item discriminates positively<br />

in that it attracts more correct responses (10)<br />

from the top 10 students than the bottom<br />

10 (8) and hence is a poor distractor; here,<br />

also, the discriminability index is 0.20, hence<br />

is a poor discriminator and is also a poor<br />

distractor. Example B is an ineffective distractor<br />

because nobody was included from either group.<br />

Example C is an effective distractor because<br />

it includes far more students from the bottom<br />

10 students (10) than the higher group (2).<br />

However, in this case any ambiguities must be<br />

ruled out before the discriminating power can be<br />

improved.<br />

Distractors are the stuff of multiple choice<br />

items, where incorrect alternatives are offered, and<br />

students have to select the correct alternatives.<br />

Here a simple frequency count of the number<br />

of times a particular alternative is selected will<br />

provide information on the effectiveness of the<br />

distractor: if it is selected many times then it is<br />

working effectively; if it is seldom or never selected<br />

then it is not working effectively and it should be<br />

replaced.<br />

If we wish to calculate the item difficulty of a test,<br />

we can use the following formula:<br />

A<br />

N × 100<br />

where<br />

A = the number of students who answered the<br />

item correctly;<br />

N = the total number of students who attempted<br />

the item.<br />

Hence if 12 students out of a class of 20 answered<br />

the item correctly, then the formula would work<br />

out thus:<br />

12<br />

× 100 = 60 per cent<br />

20<br />

The maximum index of difficulty is 100 per cent.<br />

Items falling below 33 per cent and above 67<br />

per cent are likely to be too difficult and too<br />

easy respectively. It would appear, then, that<br />

this item would be appropriate to use in a test.<br />

Here, again, whether the researcher uses an item<br />

with an index of difficulty below or above the<br />

cut-off points is a matter of judgement. In a normreferenced<br />

test the item difficulty should be around<br />

50 per cent (Frisbie 1981). For further discussion<br />

of item difficulty see Linn (1993) and Hanna<br />

(1993).<br />

Given that the researcher can know the degree<br />

of item discriminability and difficulty only once the<br />

test has been undertaken, there is an unavoidable<br />

Chapter 19

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

Saved successfully!

Ooh no, something went wrong!