Consumers satisfaction of attributes in online product design
Part IV: ANALYSIS AND RESULTS
Yüklə 1,25 Mb.
|
- Bu səhifədəki naviqasiya:
- 4.1 Data preparation
- 4.2 Analysis
Part IV: ANALYSIS AND RESULTSThis chapter reports the results of the survey conducted to test the research questions of the present study and the interpretation of the results. First the overall data preparation will be discussed followed by the analysis and leading up to the results. 4.1 Data preparationThe number of respondents who corresponded to the request to participate in the survey was 162. However, the majority of answers were missing in 60 questionnaires. As the number of the missing values was quite large and the number of the unfinished questionnaires was relatively large, these participations were not taken into account. Furthermore to composite together multiple items, all the items need to be “in the same direction”. This means that indicating a higher (or lower) response each scale must correspond conceptually to answering higher (or lower) on the other items you want to composite together. Underlining to this regulation, two questions that were negatively stated reversed. To be more specific, the questions “It is difficult to find links or menus on the website” and “Using the toolkit required little work” are reversed in a positive manner. Furthermore, the data from the questionnaire is based on two different scales, whereby the intervals of 5 point-likert are recoded into 7-point likert for a better interpretation of the results. Although, the study was conducted in the Netherlands, the questionnaire was online and therefore sanded to people of different nationalities. As mentioned above, participants should have had internet access, which implies a level of familiarization with technology. Finally, the average age of the sample was 28 years old. 4.2 AnalysisThe results were analyzed by means of conducting factor analyses and multiple regression analyses. The factor analysis refers to a collection of statistical methods for reducing correlational data into a smaller number of dimensions. Therefore first the factor analysis is conducted to understand the structure of a set of variables. The objective was to obtain results that could be better interpreted and subsequently used for further analysis. By using factor analysis, the research variables will be clustered into factors and their ability to measure each factor will be verified. The aim of this study is to examine the relationships between those dimensions and to confirm or reject the hypothesis of the conceptual framework. First, all the 39 variables were included in the present research in order to measure the respondents’ evaluation and their satisfaction with the process and design of the online co- designing. However, an extensive examination of the data showed inconsistency which made the results difficult to interpret. After careful examination of the factor results, a total of 28 variables were retained for further analysis. Second, after conducting several factor analyses and refining the data, the results obtained showed clear loadings of the attributes of online co-creation on each factor. Based on the factors obtained according to the specific item they were measuring and based on the literature review, the decision was made to rename the factors, in specifically “Web Design”, “Web Navigability”, “Transaction Capability” for website attributes, “Complexity”, “Enjoyment”, “Control” for toolkit attributes, and “Process Satisfaction” and the “Product Satisfaction”. Furthermore, in order to determine the internal consistency of the different scales, the Cronbach’s alpha was computed for each scale separately. Table 1 is presenting the results of factor analysis and an overview of Cronbach’s alpha indicating the internal consistency of the different scales. The result of the KMO indicates the appropriateness of the factor analysis. To be more specific, KMO indicates that patterns of correlations are relatively compact and factor analysis should yield distinct and reliable factors. However one of the simplest ways to estimate factor scores for each individual involves summing raw scores corresponding to all items loading on a factor (Comrey & Lee 1992). Hence, the scores estimated by the factor analysis for the different factors is replaced by the average mean of the loaded factors and will be used for further analysis, which may allow for easier interpretation. Also, average scores are useful to foster comparisons across factors when there are differing numbers of items per factor. Furthermore a few variables were removed as they were loading almost equal on two factors. The factor analysis was tested again and 7 factors were extracted this time. Note that product satisfaction and process satisfaction are extracted into two factors because of the high correlation. As indicated by the reliability tests, all the dimensions are reliable which gives us the opportunity to continue with the regression analysis in order to test the formed hypotheses.
Similar results showed in table 1 were obtained also for the items measuring process and product satisfaction. As mentioned, considering that these variables are highly correlating with each other, the decision was made to extract a separate factor analysis for product and process satisfaction, which is presented in table 2.
Yüklə 1,25 Mb. Dostları ilə paylaş:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||