05.12.2016 Views

Is headspace making a difference to young people’s lives?

Evaluation-of-headspace-program

Evaluation-of-headspace-program

SHOW MORE
SHOW LESS

Create successful ePaper yourself

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

Appendix C<br />

6 June 2014. The ‘no treatment’ group is drawn from the comparison surveys and comprises <strong>young</strong><br />

people who have not sought any substantial <strong>headspace</strong>s treatment and who also did not seek any<br />

treatment from any other health professionals between the two survey waves. The ‘other treatment’<br />

group comprises of <strong>young</strong> persons within the comparison surveys and includes all those who sought<br />

support from a health professional(s) between Waves 1 and 2. A key limitation of this group is that<br />

no information about the dose of the other treatment is known. Young people within the comparison<br />

surveys that sought a substantive <strong>headspace</strong> treatment between Wave 1 and Wave 2 were excluded<br />

from the analysis. This constitutes 90 observations.<br />

Steps were also taken <strong>to</strong> match the <strong>headspace</strong> survey intervention group <strong>to</strong> administrative (hCSA)<br />

data in order <strong>to</strong> assess the timing of the survey collection against the period of treatment at a<br />

<strong>headspace</strong> centre. This process uncovered 340 observations (32.3%) that had completed their<br />

<strong>headspace</strong> treatment prior <strong>to</strong> the Wave 1 collection date. These observations were excluded from the<br />

analysis.<br />

Difference-in-<strong>difference</strong> Method<br />

The objective of the <strong>difference</strong>-in-<strong>difference</strong> (DID) method is <strong>to</strong> assess the changes in <strong>young</strong><br />

<strong>people’s</strong> mental health, physical health, drug and alcohol use and social inclusion outcomes after<br />

using <strong>headspace</strong> services relative <strong>to</strong> other comparable <strong>young</strong> people that did not receive treatment at<br />

<strong>headspace</strong>.<br />

A <strong>difference</strong>-in-<strong>difference</strong> approach has been applied, using survey data for <strong>headspace</strong> treatment<br />

and comparison groups that have been matched on a number of different fac<strong>to</strong>rs. Differencein-<strong>difference</strong>s<br />

estimates are defined as the <strong>difference</strong> in the average outcome in the <strong>headspace</strong><br />

treatment group at two points in data collection, that is wave 1 and wave 2, minus the <strong>difference</strong> in<br />

the average outcome in the matched comparison group. Comparison groups have been matched<br />

using a propensity score matching (PSM) technique and have been divided in<strong>to</strong> groups that have<br />

received no mental health treatment between the survey waves and those that have received<br />

treatment other than <strong>headspace</strong>. Both are described further below.<br />

The <strong>difference</strong>s in outcomes for <strong>headspace</strong> clients, for matched samples drawn from the<br />

no-treatment and other-treatment comparison groups, and for the <strong>difference</strong> in <strong>difference</strong>s between<br />

<strong>headspace</strong> and either of these two matched comparison groups, are tested for their statistical<br />

significance and using effect size measures.<br />

Statistical significance is judged using an orthodox t-test for the mean <strong>difference</strong>s between the<br />

<strong>headspace</strong> treatment group and a propensity-score matched comparison group, and for the<br />

<strong>difference</strong> in the <strong>difference</strong>s between <strong>headspace</strong> and comparison groups.<br />

Effect sizes can be expressed in a number of ways, with Cohen’s d commonly reported as a<br />

standard indica<strong>to</strong>r in clinical evaluation. The Cohen effect size measure presents a standardised<br />

<strong>difference</strong> in means across the course of an intervention (ie. the ratio of mean <strong>difference</strong> <strong>to</strong> a pooled<br />

standard deviation measure). In this regard it relates <strong>to</strong> classical measures of the significance of<br />

mean <strong>difference</strong>s. It is common practice <strong>to</strong> interpret effect sizes according <strong>to</strong> the magnitude of the<br />

index – the normal ‘rule of thumb’ is <strong>to</strong> consider effects as small (0.2) through medium (0.5) <strong>to</strong> large<br />

(0.8) when relating the size of the effects.<br />

Propensity Score Matching<br />

A complexity of the DID methodology stems from the fact that those in <strong>headspace</strong> treatment group<br />

are often dissimilar in their observed and unobserved characteristics when compared <strong>to</strong> <strong>young</strong><br />

people within the comparison surveys. In order <strong>to</strong> mitigate these <strong>difference</strong>s, a propensity score<br />

matching technique has been applied <strong>to</strong> the datasets.<br />

The implementation of propensity score methods requires that a set of characteristics are chosen as<br />

the basis for alignment of the <strong>headspace</strong> treatment with a matched comparison group. It would seem<br />

appropriate <strong>to</strong> match simultaneously on a large number of fac<strong>to</strong>rs, however, the match on each single<br />

characteristic becomes less precise as the number of matching characteristics increases. For the<br />

purpose of this evaluation, four characteristics were chosen - age, gender, K10 score and the number<br />

of non-functioning days ‘out of role’ <strong>to</strong> source a match between the <strong>headspace</strong> and comparison<br />

Social Policy Research Centre 2015<br />

<strong>headspace</strong> Evaluation Final Report<br />

178

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

Saved successfully!

Ooh no, something went wrong!