FUNDED BYPARTNERS
               
Impact of COnstructing Non-motorised Networks and Evaluating Changes in Travel
register | login
Home | work areas | WP2: evaluation framework and measurement tools | produce measurement tools

Task: produce measurement tools

You need to be logged in to edit this content.

The second main task is to produce two principal measurement tools:

  • a longer postal questionnaire, mainly intended for use in researching the effects of interventions at the community or population level, and
  • a shorter user intercept questionnaire, mainly intended for distribution by hand to pedestrians and cyclists using a particular route or facility as part of the local evaluation of specific initiatives (such user intercept questionnaires are already used by Sustrans, e.g. to assess the uptake of new sections of the National Cycle Network).

In order to do this, we are collating instruments currently available for measuring travel behaviour and physical activity and for imputing carbon emissions and critique their fitness for purpose in terms of their measurement properties, their test-retest reliability, their concurrent validity relative to objective measures of activity, movement and energy consumption (e.g. pedometers, cycle computers, accelerometers, GPS devices, odometer readings and car fuel purchase receipts), and their face validity for application in the UK.

We are also assessing the predictive validity of routinely collected data (such as traffic counts and automated bicycle counters) for use as a potential proxy indicator of the population impact of infrastructure interventions on travel and physical activity behaviour as an alternative to collecting survey data. We are drawing on the expertise and previous work of the investigators and their institutions, e.g. on integrated emissions profiling at the personal level (Brand), on the development and validation of new physical activity questionnaires such as GPAQ and IPAQ (Bull) and EPAQ and RPAQ (Cambridge) and on the use of accelerometers (Cooper) and personal GPS (Preston). We are also incorporating the emerging findings from the ALPHA project, a European review of methods for assessing physical activity currently in progress (Bull, Rutter).

As well as measuring the behaviours and primary outcomes of interest, the longer questionnaire also includes items to assess characteristics of the intervention itself, the context of the intervention, and putative mediating factors. These items are being selected on the basis of a collation and synthesis of currently available items and the development of an interdisciplinary consensus as to which are most fit for purpose in this project, again drawing on the investigators' links with initiatives such as the International Physical Activity and Environment Network (IPEN) (Bull, Cooper, Mutrie, Ogilvie).

Working closely with the Sustrans Research and Monitoring Unit, with which some members of the consortium have already developed strong links, we are then piloting the feasibility and acceptability of both questionnaires and assess their test-retest reliability and concurrent validity relative to objective measures before deploying them in WPs 3 to 5. This will involve a total sample of around 600 usable responses in two waves.

Feasibility study update (January 2010)

The feasibility study was carried out during the autumn of 2009. It was designed to assess the completeness of responses, to compare the response rates by survey strategies and to predict the main survey results.

Our primary hypotheses in the feasibility design were:

  1. shorter questionnaire has better response rate
  2. personalised questionnaire has better response rate
  3. higher socio-economic status areas have higher the response rate and
  4. the reminder package elicits more responses than the reminder cards.

All these hypotheses were tested quantitatively and qualitatively. In general, the shorter, the personalised and the reminder package led significantly or marginally better responses than their own counterparts and the older, well-educated and less-deprived were more willing to respond to the postal survey. Also we achieved acceptable response quality.


by anonymous on 26 March 2010, 12 PM :
Yena Song, Southampton, 19-03-2010
site by Christian Brand, ECI