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.

VALIDITY AND RELIABILITY IN QUESTIONNAIRES 157<br />

<br />

<br />

identified under the section on ‘Threats to<br />

internal validity’.<br />

Invalidity or unreliability of instruments: The<br />

use of instruments which yield data in which<br />

confidence cannot be placed (see below on<br />

tests).<br />

Ecological validity, anditspartner,theextent<br />

to which behaviour observed in one context<br />

can be generalized to another: Hammersley<br />

and Atkinson (1983: 10) comment on the<br />

serious problems that surround attempts to<br />

relating inferences from responses gained under<br />

experimental conditions, or from interviews, to<br />

everyday life.<br />

By way of summary, we have seen that an<br />

experiment can be said to be internally valid<br />

to the extent that, within its own confines, its<br />

results are credible (Pilliner 1973); but for those<br />

results to be useful, they must be generalizable<br />

beyond the confines of the particular experiment.<br />

In a word, they must be externally valid<br />

also: see also Morrison (2001b) for a critique<br />

of randomized controlled experiments and the<br />

problems of generalizability. Pilliner (1973) points<br />

to a lopsided relationship between internal and<br />

external validity. Without internal validity an<br />

experiment cannot possibly be externally valid.<br />

But the converse does not necessarily follow; an<br />

internally valid experiment may or may not have<br />

external validity. Thus, the most carefully designed<br />

experiment involving a sample of Welsh-speaking<br />

children is not necessarily generalizable to a target<br />

population which includes non-Welsh-speaking<br />

subjects.<br />

It follows, then, that the way to good<br />

experimentation in schools, or indeed any other<br />

organizational setting, lies in maximizing both<br />

internal and external validity.<br />

Validity and reliability in questionnaires<br />

Validity of postal questionnaires can be seen<br />

from two viewpoints (Belson l986). First, whether<br />

respondents who complete questionnaires do<br />

so accurately, honestly and correctly; and<br />

second, whether those who fail to return their<br />

questionnaires would have given the same<br />

distribution of answers as did the returnees. The<br />

question of accuracy can be checked by means<br />

of the intensive interview method, a technique<br />

consisting of twelve principal tactics that include<br />

familiarization, temporal reconstruction, probing<br />

and challenging. The interested reader should<br />

consult Belson (1986: 35-8).<br />

The problem of non-response – the issue of<br />

‘volunteer bias’ as Belson (1986) calls it – can,<br />

in part, be checked on and controlled for,<br />

particularly when the postal questionnaire is sent<br />

out on a continuous basis. It involves followup<br />

contact with non-respondents by means of<br />

interviewers trained to secure interviews with<br />

such people. A comparison is then made between<br />

the replies of respondents and non-respondents.<br />

Further, Hudson and Miller (1997) suggest several<br />

strategies for maximizing the response rate to<br />

postal questionnaires (and, thereby, to increase<br />

reliability). They involve:<br />

<br />

<br />

<br />

<br />

<br />

<br />

<br />

<br />

<br />

including stamped addressed envelopes<br />

organizing multiple rounds of follow-up to<br />

request returns (maybe up to three follow-ups)<br />

stressing the importance and benefits of the<br />

questionnaire<br />

stressing the importance of, and benefits to,<br />

the client group being targeted (particularly if<br />

it is a minority group that is struggling to have<br />

avoice)<br />

providing interim data from returns to nonreturners<br />

to involve and engage them in the<br />

research<br />

checking addresses and changing them if<br />

necessary<br />

following up questionnaires with a personal<br />

telephone call<br />

tailoring follow-up requests to individuals<br />

(with indications to them that they are<br />

personally known and/or important to the<br />

research – including providing respondents<br />

with clues by giving some personal information<br />

to show that they are known) rather than<br />

blanket generalized letters<br />

detailing features of the questionnaire itself<br />

(ease of completion, time to be spent,<br />

Chapter 6

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

Saved successfully!

Ooh no, something went wrong!