The Online Tool for Ad Hoc Questionings via Internet.
|
|
|
| |
Conjoint Analyses at adenquire.net
Conjoint Analyses provide conclusions on seperate object attributes (in particular partial utility values) based on aggregate evaluations of the objects. For that, all relevant factors of an object and all relevant factor values will be implicated into a survey. Mostly, an almost unmanageable number of combinations of factor values will arise. However, to compare the combinations and therefore to estimate the partial utilities of the factors and their values, a judgement sampling by means of an orthogonal main-effect plan (OMEP) can select only a few combinations that will be sufficient for a survey. For conjoint-analytical calculations afterwards, only the so-called orthogonal condition must be met (every factor value of the selected combinations may not be correlated with any other factor value).
The Tool adenquire.net enables conjoint analyses with a maximum of 5 relevant factors and each with a maximum of 5 relevant values. By using an OMEP, there will be selected a minimum number of combinations (depending on the number of factors and factor values up to 25 combinations) that will be described on automatically generated cards. These cards have to be arranged into a ranking by the respondents according to their preferences. In addition, the respondents can also be asked for (metrical) ratings as well as for individual limits as zero points on the utility scale within the specific kind that is called limit conjoint analysis.
So, within adenquire.net there will be realised no conventionally used adaptive conjoint analysis. The adaptive method, that was developed at a time when computer screens still did not display graphical user interfaces, can barely be used for newer methodical developments. To conduct online limit conjoint analyses (also multi-stage and hierarchically individualised), a specific software support for the so-called profile method was programmed and integrated into adenquire.net.
The afterwards displayed example applies to a concrete article (handheld). This exemplary product, here, has four relevant factors each with three, more or less imaginary values. From the 81 (=3x3x3x3) possible combinations of factor values 9 (in this case the minimum) were selected (so-called reduced design). First of all, this 9 combinations have to be ranked by the respondent. In addition, this example asks for an individual limit. And coming in second, the 9 combinations displayed in the previously individually arranged ranking order should be rated by the respondent, non-contrarily to the ranking order. In place of the limit that was set before, the so-called limit card appears. (The example is completely predetermined and solely allows choosing a colouring [at the bottom of this page].)
Please note that the conjoint analysis is only available within the PremiumVersion 50 and the optional asking for ratings requires the use of a second question.
A special thanks goes to the marketing chair (held by Prof M. Voeth) at Hohenheim University (in Stuttgart, Germany) for supporting the development of the conjoint analysis question types.
|
|  |
|
|
