Data collection by traditional computerized methods, as home interviews using portable computer, has become very expensive. During 2002 the Institute of Transport Economics in Norway carried out several Stated Preference surveys using Internet software. This has proved to be less expensive and have, in general, the same advantages as other computerized interviewing methods. Using computers in Stated Preference surveys provides a great advantage in the quality of the data collections compared with manual methods. Computer based surveys allow the design to be customized to each respondent and hence improved data quality.
In Norway a large proportion has access to Internet at home or at work, and the Internet skills are in general good enough to answer an Internet-based Stated Preference survey.
The two main Internet-based surveys during 2002 were “Travellers’ valuation of traffic information based on work journey by car” and ”Better Public Transport 2002”. The two studies are similar in design and length.
The first study looked at drivers’ valuation of reduced travel time variability and delay, new forms of information in order to reduce travel time variability and delay, and the value of such information.
The second study looked at passengers’ valuation of various quality aspects of public transport in the Oslo region. The motivation was to find indicators of passengers’ preferences for alternative investments, and to compare the findings with a similar survey that was carried out in 1992. In both studies the respondents could choose to answer on the Internet or by pen-and-paper, but the respondents were encouraged to use Internet.
The recruiting was carried out differently in the two studies. While the recruiting in the first study was accomplished by delivering cards through the car-windows in the morning jam, the other study recruited respondents randomly from a national register. Comparison of the results indicates that in addition to the subject of the study, the recruiting method affects how many answers we get on Internet. The recruiting method affects the results due to the possibility to give the respondents a reminder, response rate, the percentage answering by using Internet compared to pen-and-paper, age, sex, occupation and distribution of the responses during the day and over the week.
Our experience is that the results using the Internet are more reliable than home interview using portable computer.
Sawtooth Software has developed the software for data collection used in our surveys. It contains a module to write questions, a password-module and an on-line data management module. Our surveys are based on Adaptive Conjoint Analysis for Web (ACA/Web) and Advanced Choice-Based Conjoint for Web (CBC/Web). SPSS and Alogit are used for analysis purposes. In addition to the two studies mentioned earlier we are preparing a third study this spring in Sweden. Our paper will present experience from these studies and focus on:
In what sort of studies would we derive advantage from using Internet?
In what sort of studies would we derive advantage from this particular software?
Do Internet-based studies require special design?
How do we get a representative sample using Internet?
How do we to get the desired respondents using Internet?
| Back to: Top | Programme | Page last updated on 31 August, 2003 |