About the Survey
Taiwan Communication Survey
The Taiwan Communication Survey collects and compiles data on the Taiwanese general population’s media consumption as well as the patterns of the consumption. Using rigorous sampling methodologies, the Survey interviews 2,000 people in Taiwan annually to investigate how they consume media, providing the only instant and representative survey of this kind in Taiwan.
The work of surveying the general population’s media consumption in Taiwan began in 1984, when the National Science Council assigned the Taiwan Social Change Survey Project to Academia Sinica, which formed a “mass communication group” to carry out island-wide surveys once every five years. With the advent of new technology and the change of global media environment, the “mass communication group” officially broke away from the Taiwan Social Change Survey Project and became an independent project known as Taiwan Communication Survey, or TCS. Scholars from related fields were gathered to form a committee, and surveys were conducted annually instead of once every five years so as to better observe the general population’s media consumption behavior and to make the data more instant and referable.
To better understand how technological development affects and changes media consumption behavior, Taiwan Communication Survey not only conducts annual surveys on media consumption behavior, but also focuses on different sub-topics each year. With these long-term studies, the Survey can create a more comprehensive and thematic database, helping researchers analyze and theorize Taiwan’s current situation and future trends, as well as providing the government with a basis to formulate future communication-related policies.
The Taiwan Communication Survey Project, carried out by a committee of communication scholars, includes a six-year study (from October 2011 to September 2017) during its first phase, and each year the study focuses on a different sub-topic.
The first sub-topic of the first phase (2012) is a comprehensive survey on mass communication behaviors, targeting news, advertisements, politics, public health and consumption. At the same time, the study also explores the general population’s interpersonal interactions and interpersonal mobilizations in this age of new media (the survey lasted from July to September 2012, and the data has been released in July 2013).
The second sub-topic of the first phase (2013) focus on the Internet, investigating the general population’s Internet usage and behavior, especially with regard to social networking sites, dual media, digital entertainment, e-commerce, online curation, and Internet literacy. Furthermore, the study will utilize network detection software to objectively observe web users’ online behavior (the survey will be carried out from July to September 2013, and the data will be released in July 2014).
The third sub-topic of the first phase (2014) is center on entertainment and entertainment-related media behavior. As minors constitute a large part of the population that uses entertainment media, the age limit for the respondents in this part of the study will be lowered to nine, and two separate versions will be prepared for the adults and the minors (the survey will be carried out from July to September 2014, and the data will be released in September 2015).
The fourth sub-topic of the first phase (2015) focus on the political communication behaviors, political and electoral opinions, attitudes, and cognition, and political and electoral behaviors. (the survey lasted from July to September 2015, and the data has been released in July 2016).
The fifth sub-topic of the first phase (2016) focus on risk and disaster communication behaviors, risk and disaster attitudes and cognition, and risk and disaster behaviors. (the survey lasted from July to September 2016, and the data has been released in July 2017).
The sixth sub-topic of the first phase (2017) focus on utility of new media (e.g., social interaction, information access, entertainment) and impacts of Internet use (e.g., addiction, cyberbullying and problematic use). (the survey will be carried out from July to September 2017, and the data will be released in July 2018).
In terms of respondent sampling, the population of concern includes Republic of China citizens in Taiwan over the age of 18; Taiwan’s census data is used as the sampling frame, and probability-proportional-to-size (PPS) sampling is employed as the sampling method. The population is stratified based on factors such as population density, level of education, percentage of seniors aged over 65, percentage of population aged between 15 and 64, percentage of population working in the industry sector, and percentage of population working in the business sector, dividing Taiwan into six strata.
In terms of the surveying process, interviews were conducted by a well-trained survey team, led by professors with years of experience in the field. First, all supervisory staff members underwent training, so that each member would be clear on his or her work responsibilities and content. Interviewers also received training, in which they were taught interviewing techniques as well as other things to bear in mind. The training emphasized on questionnaire reading and interview practices, so that interviewers could familiarize themselves with the questionnaires and be prepared for different situations during live interviews. Trial interviews were conducted prior to the real survey to identify potential problems, and evaluations were made to revise the questionnaires and fix the issues. Thus, the quality of the interviewers was ensured for this survey, and the credibility of this survey was also optimized in the process.
In addition to innovating and expanding the questionnaires, the six-year survey broke away from traditions and conventions by developing its very own CAPI software, using tablet computers to carry out the interviews. This move was adopted because of the intricate complexity of mass communication behaviors. When respondents answer questions, their response to certain questions may affect whether or not subsequent questions need to be answered. With traditional paper questionnaires, skipping questions may lead to confusion and errors, causing sample attrition. However, CAPI can minimize potential errors by automatically skipping unneeded questions, as is proven by the interviews, while its data uploading function can also expedite the data entry and checking processes after interviews. Moreover, data collected will also be coupled with online tracking to better investigate the general population’s online media consumption behavior.