30.06.2013 Views

Sick and Tired: Understanding and Managing Sleep Difficulties in ...

Sick and Tired: Understanding and Managing Sleep Difficulties in ...

Sick and Tired: Understanding and Managing Sleep Difficulties in ...

SHOW MORE
SHOW LESS

Create successful ePaper yourself

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

As previously described on page 61, the questionnaire data was scored accord<strong>in</strong>g to the<br />

<strong>in</strong>dividual questionnaire <strong>in</strong>structions <strong>and</strong> the result<strong>in</strong>g variables were checked for<br />

skewness <strong>and</strong> kurtosis. Transformation of skewed variables aims to normalise the<br />

distribution of the data <strong>in</strong> order to meet the assumptions of parametric statistics (Norris<br />

& Aroian, 2004). There are several different approaches for transform<strong>in</strong>g data,<br />

dependent on the degree <strong>and</strong> direction of the skew. Due to the need to use different<br />

transformation approaches for different distributions of data, this can cause difficulties<br />

<strong>in</strong> apply<strong>in</strong>g transformations, as all items on a given scale need to be transformed us<strong>in</strong>g<br />

the same approach to enable the comparison of like with like, <strong>and</strong> one approach may not<br />

be suitable for all items with<strong>in</strong> a scale. Consequently, transformation of skewed data is<br />

not always recommended (Norris & Aroian, 2004). As the data were considered to be<br />

ord<strong>in</strong>al <strong>in</strong> this thesis (as discussed <strong>in</strong> Chapter Four) non-parametric statistics were used<br />

to analyse the data <strong>in</strong> this study. Non-parametric tests do not require data to be normally<br />

distributed <strong>and</strong> therefore the skewness of the study variables were not corrected.<br />

To account for multiple comparisons, f<strong>in</strong>d<strong>in</strong>gs were only considered to be significant at<br />

the p

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

Saved successfully!

Ooh no, something went wrong!