Systematic review and quantitative meta-analysis [Elektronische Ressource] : methodological foundations and practical applications in the domain of transport policy / vorgelegt von Guido Möser

Sujets

• Soziologie

Informations

• Publié par: justus-liebig-universitat_giessen
• Publié le: 01 janvier 2007
• Nombre de lectures: 24

Extrait

SYSTEMATIC REVIEW AND QUANTITATIVE META-ANALYSIS:
METHODOLOGICAL FOUNDATIONS AND PRACTICAL
APPLICATIONS IN THE DOMAIN OF TRANSPORT POLICY





Dissertation zur Erlangung des Doktorgrades (Dr. rer. soc.)
des Fachbereichs Gesellschaftswissenschaften
der Justus-Liebig-Universität Giessen




Vorgelegt von

Guido Möser
aus Gießen

2006
Supervisor: Prof. Dr. Peter Schmidt
2 SYSTEMATIC REVIEW AND QUANTITATIVE META-ANALYSIS: METHODOLOGICAL
FOUNDATIONS AND PRACTICAL APPLICATIONS IN THE DOMAIN OF TRANSPORT POLICY


Content

1) Introduction ......................................................................................................6
Acknowledgements........................................................................................21


2) Meta-analysis: An alternative to narrative reviews for
synthesising social science research? (Zusammen mit
Prof. Dr. Peter Schmidt) ..................................................................................23


3) Are ‘Soft’ Policy Measures Effective in Reducing Peoples’ Car
Use? A Meta-Analytical Review of Research Evidence
(Zusammen mit PD Dr. Sebastian Bamberg)...................................................60


4) Are Work Travel Plans Effective? – Systematic Review and
Meta Analysis in the Transport Policy Domain
(In Koautorenschaft mit PD Dr. Sebastian Bamberg ......................................107
3

5) Twenty years after Hines, Hungerford, and Tomera: A new
meta-analysis of determinants of pro-environmental behavior
(In Koautorenschaft mit PD Dr. Sebastian Bamberg).....................................154


6) Conclusions..................................................................................................187


7) References for the Introduction and the Conclusion Sections................202


8) Appendix
I) A meta-analysis of the impact of new yellow school buses
on pupils transport to school..............................................................206
II) Erklärung ..............................................................................................209
III) Spezifizierung der Koautorenschaft...................................................210



4 5 1) Introduction

In this dissertation I will try to examine if quantitative methods in synthesising
single study findings are helpful in the domain of transport policy. An introductory
example deals with the problem:

Suppose there is a community called A-City, which is confronted by a lot of traffic. Cars are
especially causing air and noise pollution, street damage, traffic jams etc. A-City-government is
interested in reducing the number of people travelling by car. A-City is already running public
busses, trains and a school travel system for pupils. The main goal of A-City government is to
reduce people travelling by car. Due to a limited budget, A-City government has to work cost
effectively. In this example, A-City government is right at the starting point of investigating this
problem.

Scientists, politicians, policy makers and transportation specialists have wide-
ranging information needs and limited budgets. For that reason, they need reliable
information on the effectiveness of a large number of different kinds of
interventions in the domain of transportation. Moreover, many programmes on
national and European policies focus on the problem of an appropriately handling
transport policies. For example, “the European Commission’s objective for the next
ten years is to refocus Europe’s transport policy on the demands and needs of its
citizens (European Commission, 2001).” Measures recommended by the
European Commission are for example improving road safety or preventing
congestion.

A-City held a conference, inviting experts and staff from other local governments who had
already tried to reduce single occupied vehicle usage (cars and motorcycles). At this conference
6 different kinds of travel interventions were discussed. After the conference, A-City’s Government
was not satisfied because some of the invited participants reported strong difficulties when
implementing different kinds of travel interventions. The implementation of flawed strategies led to
massive expenditures and the main goal of reducing car use was not attained. Therefore, A-City’s
government consulted different sources in order to find out more about the effectiveness of
different kinds of policy measures to reduce travelling by car. The main sources, travel policy
journals, were reviewed. Since the number of articles reporting different results for the same kind of
intervention in reducing car use was very high, A-City’s government was dissatisfied again.

Taking into consideration the economic purposes and the limited budget of A-
City Government, it is necessary for them to implement the most effective car
reducing intervention programmes. As a first step, an investigation of other
programmes and a literature review will be helpful. If the city government does not
obtain satisfying results, a professional researcher should be hired.

A-City government hired a scientific researcher. The researcher collected all literature about
travel interventions to reduce car use published so far. Finally, after reading the collected literature
carefully, he summarised the results in a report. This report showed no concrete statistics on
effectiveness but his own scientific interpretation of the studies.

It seems, that the scientific researcher used techniques of the so called
‘narrative reviews’ in the example above. In a narrative review, a researcher
collects all the relevant information on a special subject to summarise the
information for other readers interested in the field. The researcher has to choose
which information is relevant and interesting for other readers. This selection
process may result in difficulties: “Narrative reviews do not reveal how the
decisions were made about relevance of studies and the validity of the included
7 studies.” (Colins & Fauser, 2004, p. 103). Thus in reporting and summarising the
results of synthesising the single study results, “[d]ifferent reviewers were reaching
different conclusions from the same research base and the findings reported often
had more to do with the speciality of the reviewer than with the underlying
evidence.” (Colins & Fauser, 2004, p. 104).

The next problem one may assume from the example above is that although
narrative reviews are an old and established way to synthesise single study
findings, they fail to report concrete statistical parameters on the effectiveness of
the results of all studies together. How strong is the overall direction and
magnitude of interventions in reducing speed limits to car use? Do interventions
lead to significant changes (a reduction of five to ten percent)? Is there an effect at
all?

Fortunately, a member of A-City government met an old friend who was a researcher in an
institute. This institute was interested in investigating the effectiveness of different kinds of travel
intervention programmes. They discussed the problems of A-City government exchanging
information on the effectiveness of different travel intervention programmes. The researcher
informed his old friend from the A-City government about newer approaches to report statistical
parameters on the effectiveness of different kinds of travel intervention programmes. The statistical
parameters had been synthesised out of single study results. These approaches used quantitative
statistical techniques to solve the problem of summarising single study findings.

In the example above, the researcher talked about quantitative methods to
summarise research findings from single studies. These methods are the so called
8 systematic review and meta-analysis. Both methods offer ways to synthesise the
results from a large body of empirical studies in a systematic and objective way.

Systematic review and meta-analysis could help the A-City government to
obtain a systematic overview of different kinds of travel interventions through
synthesising the results from all studies at hand.

Systematic reviews are a further development of the narrative reviews. A
systematic review is a better way of summarising research evidence than the
narrative review: “Systematic reviews are like scientific investigations in
themselves, using pre-planned methods and an assembly of original studies that
meet their criteria as ‘subjects’. They synthesise the results of an assembly of
primary investigations using strategies that limit bias and random error.”
(Cochrane Muscoloskeletal Group, 2006). In other words, a systematic review is
any type of review using strategies to avoid bias and including a material and
methods section. An example of a strategy used to avoid bias is to identify all
relevant studies published since 1985. In the material section of a systematic
review all used studies are mentioned. In the method section, methods to
summarise the res