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.

508 QUANTITATIVE DATA ANALYSIS<br />

<br />

<br />

<br />

<br />

Most respondents thought that the course was<br />

‘a little’ too hard (with a response number of<br />

98, i.e. 51.3 percent); the modal score is that<br />

category or score which is given by the highest<br />

number of respondents.<br />

The results were skewed, with only 10.5 per<br />

cent being in the categories ‘quite a lot’ and ‘a<br />

very great deal’.<br />

More people thought that the course was ‘not<br />

at all too hard’ than thought that the course<br />

was ‘quite a lot’ or ‘a very great deal’ too hard.<br />

Overall the course appears to have been slightly<br />

too difficult but not much more.<br />

Let us imagine that we wished to explore this<br />

piece of datum further. We may wish to discover,<br />

for example, the voting on this item by<br />

males and females. This can be presented in a<br />

simple cross-tabulation, following the convention<br />

of placing the nominal data (male and female) in<br />

rows and the ordinal data (the 5-point scale) in<br />

the columns. A cross-tabulation is simply a presentational<br />

device, whereby one variable is presented<br />

in relation to another, with the relevant data inserted<br />

into each cell (automatically generated by<br />

software packages, such as SPSS) (Box 24.2) (see<br />

http://www.routledge.com/textbo<strong>ok</strong>s/<br />

9780415368780 – Chapter 24, file 24.5.ppt).<br />

Box 24.2 shows that, of the total sample, nearly<br />

three times more females (38.2 per cent) than<br />

males (13.1 per cent) thought that the course<br />

was ‘a little’ too hard, between two-thirds and<br />

three-quarters more females (19.9 per cent) than<br />

males (5.8 per cent) thought that the course was a<br />

‘very little’ too hard, and around three times more<br />

males (1.6 per cent) than females (0.5 per cent)<br />

thought that the course was ‘a very great deal’ too<br />

hard. However, one also has to observe that the size<br />

of the two subsamples was uneven. Around threequarters<br />

of the sample were female (73.8 per cent)<br />

and around one-quarter (26.2 per cent) was male.<br />

There are two ways to overcome the problem<br />

of uneven subsample sizes. One is to adjust<br />

the sample, in this case by multiplying up the<br />

subsample of males by an exact figure in order<br />

to make the two subsamples the same size<br />

(141/50 = 2.82). Another way is to examine the<br />

data by each row rather than by the overall totals,<br />

i.e. to examine the proportion of males voting<br />

such and such, and, separately, the proportion<br />

of females voting for the same categories of the<br />

variable (Box 24.3).<br />

If you think that these two calculations and<br />

recalculations are complicated or difficult (overallpercentaged<br />

totals and row-percentaged totals),<br />

then be reassured: many software packages, e.g.<br />

SPSS (the example used here) will do this at one<br />

keystr<strong>ok</strong>e.<br />

In this second table (Box 24.3) one can observe<br />

that:<br />

<br />

There was consistency in the voting by males<br />

and females in terms of the categories ‘a little’<br />

and ‘quite a lot’.<br />

Box 24.2<br />

Cross-tabulation by totals<br />

Sex* The course was too hard: cross-tabulation<br />

The course was too hard<br />

Not at all Very little A little Quite a lot A very great deal Total<br />

Male Count 7 11 25 4 3 50<br />

%oftotal 3.7% 5.8% 13.1% 2.1% 1.6% 26.2%<br />

Female Count 17 38 73 12 1 141<br />

%oftotal 8.9% 19.9% 38.2% 6.3% 0.5% 73.8%<br />

Total Count 24 49 98 16 4 191<br />

%oftotal 12.6% 25.7% 51.3% 8.4% 2.1% 100.0%

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

Saved successfully!

Ooh no, something went wrong!