With the rapid growth of information and communication technology (ICT) in Korea, there was a need to improve the quality of official ICT statistics. In order to do this, various factors had to be considered, such as the quality of surveying, processing, and output as well as the reputation of the statistical agency. We used PLS estimation to determine how these factors might influence customer satisfaction. Furthermore, through a comparison of associated satisfaction indices, we provided feedback to the responsible statistics agency. It appears that our model can be used as a tool for improving the quality of official ICT statistics.
Today, globalization, deregulation, and innovation are propelled by information and communication technologies (ICT) and are they are changing the economic landscape [29]. For effective decision-making, the quality of ICT statistics is important. However, rapid changes often complicate the generation of high quality official statistics. One problem is the lack of a standardized ICT classification system; also there are few experts familiar with both ICT applications and statistics.
In order to achieve high quality official statistics, Sung [30] emphasized the importance of effective operation of the statistical agency by applying TQM. Helenius and Liewendahl [9] emphasized two important components: grasping customer's needs and training staff according to them. In addition, Lee and Sohn [23], Haworth [8], and Sonnberger and Linden [28] assessed the statistical quality programs of their own country and suggested ways for improvement. According to Eurostat [5], deterioration of statistics quality is related to several sources of error: sampling, coverage, measurement, processing, and non-response. However, most approaches are based on forming check lists or arguing opinion rather than determining satisfaction of customer need. In order to manage the level of quality of, a more systematic approach is required.
In Korea, KAIT (Korea Association of Information & Telecommunication) is the responsible agency that compiles eight kinds of ICT-related official statistics about the conditions of the industry, the state of its employment, and trends in the ICT market [10], [11], [12], [13], [14], [15], [16] and [17]. However, the users have not been sure that the official statistics were accurate and reliable and thus the responsible agency became concerned with monitoring the level of statistics quality, understanding customers’ needs, and improving data quality. Our goal was to suggest a systematic method for resolving these concerns.
The quality of official statistics is dependent on the data collection procedure, data processing procedures, result reporting, etc. We used a systematic method based on SEM (the structural equation model), considering the relationships between the quality dimensions to increase the quality of official statistics.
As the ICT industry grows in Korea, governmental decision makers and corporations depend more on the availability of good ICT statistics. However, their quality had not been properly evaluated. Consequently, we considered the relationship among the quality factors for the ICT statistics and compared the levels of satisfaction of each customer group.
In our analysis, we found that survey form, processing, and output quality plus the reputation of survey institution all had an effect on customer satisfaction. Particularly, survey form and processing quality had an indirect effect while output quality and reputation of the survey institution had direct effect on satisfaction. Customer care by the survey agency had the highest direct effect on customer satisfaction, while the survey form quality had the highest indirect effect. With a better computer environment, customers were more satisfied with the web service than the report.
We thus infer that customer care is important in improving satisfaction, that improving only some of the factors will not provide better outcomes and that the structural relationship between factors is more important.
Through comparing the averages of measured variables, we have shown what is needed to provide better services for different customer groups. Our customer satisfaction index can be used to reflect different customer's needs and also as a benchmark for improving the quality of official statistics.
Our model was applied to improve the quality of official statistics on the information and technology industry in Korea. This approach could, of course, be applied to official statistics in other industrial areas: then we could measure the level of quality for each system and the resulting index could be used to compare different systems and also as a feedback mechanism for improving the quality of official statistics.
