15.12.2012 Views

Proofs - Personal Webspace for QMUL - Queen Mary, University of ...

Proofs - Personal Webspace for QMUL - Queen Mary, University of ...

Proofs - Personal Webspace for QMUL - Queen Mary, University of ...

SHOW MORE
SHOW LESS

Create successful ePaper yourself

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

our study, we composed an assessment tool in the <strong>for</strong>m <strong>of</strong> a<br />

checklist based on them. Because our research is focusing on the<br />

concept <strong>of</strong> “intercultural opening” as it moves at an abstract and<br />

cross- institutional level, tailoring different modules <strong>for</strong> different<br />

settings <strong>of</strong> institutions was deemed unnecessary. For general<br />

in<strong>for</strong>mation about the institutions in the study, we used one part<br />

<strong>of</strong> a questionnaire from a <strong>for</strong>mer study project. The preliminary<br />

assessment tool we developed included 17 different parts and<br />

17.5 pages.<br />

A further review <strong>of</strong> the instrument was essential in order<br />

to insure that the developed tool incorporated all facets <strong>of</strong><br />

“intercultural opening”, that it was applicable to every type <strong>of</strong><br />

institution later assessed and to shorten it. We used two methods<br />

combining theoretical and practical views in two subsequent<br />

steps.<br />

3.2. Web- based, two- round<br />

consensus- oriented Delphi process<br />

As the concept <strong>of</strong> “intercultural opening” is an expertdesigned<br />

concept, in a fi rst step, we used a web- based expert<br />

interview, based on a two- round consensus- oriented Delphi<br />

process [21] to review the developed instrument <strong>for</strong> missing<br />

aspects <strong>of</strong> the instrument and, if possible, to shorten it.<br />

The general idea <strong>of</strong> the Delphi approach is to collect expert<br />

opinions on a topic in successive interview rounds and to publish<br />

the anonymous results from each round to everybody in the<br />

next round. Some Delphi approaches are consensus- oriented,<br />

which means that the desired goal <strong>of</strong> the Delphi process is<br />

consensus. The number <strong>of</strong> participants in Delphi processes<br />

has varied in other studies [11]. We decided to interview 12<br />

leading German experts (see acknowledgment) with experience<br />

and/or expertises in the fi elds <strong>of</strong> immigration, improving<br />

access to care <strong>for</strong> migrants, and the concept <strong>of</strong> “intercultural<br />

opening”. These experts come from both scientifi c and practical<br />

background, which also speaks to their different views on the<br />

topic <strong>of</strong> “intercultural opening”. The participants were chosen<br />

by conducting a research <strong>for</strong> experts in these fi elds as well<br />

as by recommendations <strong>of</strong> experts [11]. They were contacted<br />

by email and asked <strong>for</strong> their participation. Only four out <strong>of</strong><br />

the twelve selected experts declined participation and were<br />

substituted by four different recommended experts from the<br />

same areas as those who had declined to participate. One expert<br />

had to be excluded after the fi rst round due to incomplete<br />

answers.<br />

3.2.1. First round <strong>of</strong> Delphi process<br />

The first round <strong>of</strong> the Delphi process via internet was<br />

conducted from October 2010 until early January 2011. A<br />

website was created displaying the preliminary assessment<br />

tool as developed to that point. In the fi rst round, three questions<br />

were used to evaluate each item in the assessment tool<br />

and its corresponding response categories. The fi rst question<br />

concerned the signifi cance <strong>of</strong> the particular items in the context<br />

<strong>of</strong> “intercultural opening”, with the response categories “very<br />

important”, “important”, “less important” and “not important<br />

S. Penka et al. / European Psychiatry 27 (2012) / supplement n°2 / S63-S69 S65<br />

at all”. The other questions asked <strong>for</strong> an assessment <strong>of</strong> the question’s<br />

wording and <strong>for</strong> comments on the questions/items and<br />

response categories in general. Overall 170 items had to be rated<br />

by the interviewed experts. After the data were collected, the<br />

results <strong>of</strong> the fi rst round were analysed.<br />

Because <strong>of</strong> our primary goal <strong>of</strong> shortening the tool, essential<br />

criterion were defi ned as being rated “important” or “very<br />

important” by every expert. Each item that was judged as either<br />

“important” or “very important” by all experts (“Md” 12) was<br />

considered consensual and essential <strong>for</strong> determining the status <strong>of</strong><br />

“intercultural opening”. Items reaching this criterion were kept<br />

in the assessment tool. Those which did not reach the criterion<br />

were presented again <strong>for</strong> evaluation in the following round.<br />

3.2.2. Results <strong>of</strong> the fi rst round <strong>of</strong> Delphi process<br />

In the first round, 62 items reached the definition <strong>of</strong><br />

consensus and were there<strong>for</strong>e approved <strong>for</strong> use in the assessment<br />

tool, thus playing no further role in the second round.<br />

108 items, on the other hand, had failed this condition, not<br />

having been judged as “important” or “very important” by all<br />

experts. However, the majority <strong>of</strong> these items were judged as<br />

less or not important by only one or two experts. 41 items, <strong>for</strong><br />

example, had been rated by only one expert as “less important”,<br />

while 13 by only one expert as “not at all important”.<br />

54 items thereby only reached Md= 10 within the fi eld <strong>of</strong><br />

“importance”. 31 items had been judged by two experts as<br />

not (very) important (Md= 9).<br />

In addition, we worked some changes into the questionnaire<br />

based on the comments given regarding the open questions.<br />

Some items were made more specifi c, as experts had told us that<br />

they could easily be misunderstood. Some response categories<br />

were added that experts had missed. Some questions were split<br />

into two, which explains the higher number <strong>of</strong> items to be judged<br />

in the second round.<br />

<strong>Pro<strong>of</strong>s</strong><br />

3.2.3. Second round <strong>of</strong> Delphi process<br />

The second round <strong>of</strong> Delphi process took place between<br />

January and February 2011. We displayed the revised questionnaire<br />

on the website. Only the 116 items and the corresponding<br />

response not found consensual in the fi rst round were included.<br />

Every participant was asked to evaluate them again. We published<br />

the ratings <strong>of</strong> each <strong>of</strong> these items in the fi rst Delphi round.<br />

In order to get new refl ections and judgement <strong>of</strong> the topic and<br />

to avoid simple repetition, we did not in<strong>for</strong>m participants about<br />

their own previous judgements [11].<br />

3.2.4. Results <strong>of</strong> the second round <strong>of</strong> Delphi process<br />

We analysed the collected data using the same criterion<br />

as in the fi rst round. However, after the second round items<br />

which did not reach the criterion <strong>of</strong> consensus were taken<br />

out <strong>of</strong> the assessment tool. 65 items reached consensus in the<br />

second round while 51 did not. Again the majority <strong>of</strong> these<br />

items were judged as less or not important by no more than

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

Saved successfully!

Ooh no, something went wrong!