learning-styles
learning-styles
learning-styles
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
Shwery (1994) also questioned aspects of the LSI:<br />
‘The instrument is still plagued by issues related to its<br />
construct validity and the lack of an a priori theoretical<br />
paradigm for its development.’<br />
Reliability<br />
Curry (1987) judged the internal reliability of the LSI<br />
and PEPS to be good, with an average of 0.63 for the<br />
LSI and 0.66 for the PEPS. Yet she did not indicate<br />
what she regarded as ‘good’ coefficients and these are<br />
normally accepted to be 0.7 or above for a sub-scale.<br />
LaMothe et al. (1991) carried out an independent study<br />
of the internal consistency reliability of the PEPS with<br />
470 nursing students. They found that only 11 of the<br />
20 scales had alpha coefficients above 0.70, with the<br />
environmental variables being the most reliable and<br />
the sociological variables the least reliable.<br />
Knapp (1994) 6 expressed concerns both about<br />
the approach to reliability in the design of the LSI<br />
and the reporting of reliability data: in particular,<br />
he criticised repeating questions in the LSI to improve<br />
its reliability. He added:<br />
No items are, in fact, repeated word for word. They<br />
are simply reworded … Such items contribute to<br />
a consistency check, and are not really concerned<br />
with reliability at all … Included in the directions<br />
on the separate answer sheet … is the incredible<br />
sentence ‘Some of the questions are repeated to help<br />
make the inventory more reliable’. If that is the only<br />
way the authors could think of to improve the reliability<br />
of the inventory, they are in real trouble!<br />
There are also concerns about the Dunns’ claims for<br />
internal consistency. For example, Shwery (1994) says:<br />
Scant evidence of reliability for scores from the LSI<br />
is provided in the manual. The authors report [that]<br />
‘research in 1988 indicated that 95 percent’ (p.30)<br />
of the 22 areas … provided internal consistency<br />
estimates of 0.60 or greater. The actual range is<br />
0.55–0.88. Internal consistency of a number of areas …<br />
was low. As such, the link between the areas and<br />
justifiably making decisions about instruction in these<br />
areas is questionable.<br />
Murray-Harvey (1994) reported that the reliability<br />
of ‘the majority’ of the PEPS elements was acceptable.<br />
However, she considered ‘tactile modality’ and<br />
‘<strong>learning</strong> in several ways’ to ‘show poor internal<br />
consistency’ (1994, 378). In order to obtain retest<br />
measures, she administered the PEPS to 251 students<br />
in 1991 and again in 1992. Environmental preferences<br />
were found to be the most stable, with coefficients<br />
of between 0.48 (‘design’) and 0.64 (‘temperature’),<br />
while sociological and emotional preferences were less<br />
so (0.30 for ‘persistence’ and 0.59 for ‘responsibility’),<br />
as might be expected from Rita Dunn’s (2001a)<br />
characterisation of these areas as more open to<br />
change. However, the physiological traits, which are<br />
supposed to be relatively stable, ranged from<br />
0.31 for a specific ‘late morning’ preference to 0.60<br />
for a general ‘time of day’ preference (Price and Dunn<br />
1997). Overall, 13 out of 20 variables exhibited poor<br />
test–retest reliability scores of below 0.51.<br />
Two separate reviews of the PEPS by Kaiser (1998)<br />
and Thaddeus (1998) for the Mental Measurements<br />
Yearbook highlighted concerns about the Dunns’<br />
interpretations of reliability. Both reviews noted the<br />
reliability coefficients of less than 0.60 for ‘motivation’,<br />
‘authority-oriented <strong>learning</strong>’, ‘<strong>learning</strong> in several ways’,<br />
‘tactile <strong>learning</strong>’ and ‘kinaesthetic <strong>learning</strong>’. Thaddeus<br />
also noted that some data was missing, such as<br />
the characteristics of the norm group to whom the<br />
test was administered.<br />
Validity<br />
Criticism was directed at a section entitled ‘reliability<br />
and validity’ in the LSI manual (Price and Dunn 1997,<br />
10). Knapp (1994) argued that ‘there is actually<br />
no mention of validity, much less any validity data’<br />
and Shwery (1994) noted that ‘the reader is referred<br />
to other studies to substantiate this claim’. These<br />
are the dissertation studies which supporters cite<br />
to ‘provide evidence of predictive validity’ (De Bello<br />
1990, 206) and which underpin the meta-analyses<br />
(Dunn et al. 1995). There were also problems in<br />
obtaining any information about validity in the PEPS<br />
(Kaiser 1998; Thaddeus 1998) and a problem with<br />
extensive lists of studies provided by the Dunns,<br />
namely that: ‘the authors expect that the validity<br />
information for the instrument can be gleaned through<br />
a specific examination of these studies.’ (Kaiser 7 1998).<br />
Kaiser also makes the point that ‘just listing the<br />
studies in which the PEPS was used does not add<br />
to its psychometric properties’.<br />
6<br />
Page numbers are not available for online Buros reports from the<br />
Mental Measurements Yearbooks. The same applies to Shwery (1994).<br />
7<br />
Page numbers are not available for online Buros reports from the Mental<br />
Measurements Yearbooks. The same applies to Thaddeus (1998).