Introduction
Most internet users in Taiwan are young adults and middle-aged people, and elderly adults are less likely to use the internet and information media. However, because of the continuous development of information technology (IT), internet usage among older adults is gradually increasing and becoming more mainstream. Because of the ageing population in Taiwan, concerns over health are on the rise, and this is directly related to the products that people eat. As health consciousness (i.e. the degree to which consumers are interested in their health and motivated to engage in preventive behaviours and health care) increases, people develop greater tendencies to consider remedies to fight the obesity epidemic, and many health campaigns have addressed this motivation (Walls et al., Reference Walls, Peeters, Loff and Crammond2009). Thus, close attention should be paid to health consciousness. The present study hypothesised that despite the gradual increase in internet usage, elderly adults are not as technically proficient as young adults or middle-aged people and sometimes cannot find relevant websites or information to clarify their doubts. Thus, elderly adults often encounter problems such as browsing difficulties when seeking information. This study initially investigated health consciousness and IT by using the technology acceptance model (TAM) as the main model combined with attitude and trust as an extension of the model.
Since 2007, vegetables, fruits and fish in markets in Taiwan have been accompanied by small green labels, which are commonly known as traceable agricultural products (TAPs). Most agricultural and fishery products marked with these labels appear on TAP counters specially set up in supermarkets and have slightly higher prices than similar products not denoted as TAPs. By inputting the traceability number on the Taiwan Agricultural and Food Traceability (TAFT) website, consumers can track back information on TAPs to know where the products come from. The information includes the production region, farmer, planting and feeding processes, harvest or slaughter periods, time of packaging and shipping, and most crucially, results of product pesticide or other drug residue detection (Council of Agriculture, 2017). This information is comparable to records of an individual's family, education and work history, and thus is termed the production and marketing traceability of agricultural products.
Internet usage is becoming more common within a wider age range in Taiwan, and thus analysing the behaviours of elderly adults with high health consciousness when they are using the TAFT website is necessary to determine the influence of health consciousness on food attitudes and purchase intention (Michaelidou and Hassan, Reference Michaelidou and Hassan2008; Tarkiainen and Sundqvist, Reference Tarkiainen and Sundqvist2009). The main objectives of the present study were to develop an integrated extensibility model incorporating the TAM and to investigate the impact of health consciousness on elderly adults’ acceptance of technology in relation to traceability information websites. The continuous development of IT has enabled agricultural product traceability systems to be integrated into information systems. Consumers can scan a product's QR code by using their mobile phones to connect to the TAFT website and access relevant information on the product in question. Moreover, consumers can access information on the TAFT website anywhere and at any time through the website's information interface. Consumers are largely influenced by their levels of health consciousness, and tend to be concerned for their wellbeing and motivated to improve or maintain their health and quality of life in order to prevent illness through engagement in healthy behaviours. Such behaviours in turn affect the attitudes that people adopt when using agricultural product traceability systems.
Literature review and hypothesis development
TAM
The TAM, originally developed by Davis (Reference Davis1989), has subsequently been modified to create a powerful and parsimonious model (Yousafzai et al., Reference Yousafzai, Foxall and Pallister2007a, Reference Yousafzai, Foxall and Pallister2007b, Reference Yousafzai, Foxall and Pallister2010). The TAM has been used extensively to investigate various aspects of IT such as e-commerce (McKechnie et al., Reference McKechnie, Winklhofer and Ennew2006; Bigné-Alcañiz et al., Reference Bigné-Alcañiz, Ruiz-Mafé, Aldás-Manzano and Sanz-Blas2008; Tong, Reference Tong2010), e-books (Read et al., Reference Read, Robertson and McQuilken2011), e-health care (Lanseng and Andreassen, Reference Lanseng and Andreassen2007), mobile devices (Huang et al., Reference Huang, Lin and Chuang2007; Kim and Garrison, Reference Kim and Garrison2009), Web 2.0 (Shin and Kim, Reference Shin and Kim2008), internet banking (Yousafzai et al., Reference Yousafzai, Foxall and Pallister2010) and online group buying (Wang and Chou, Reference Wang and Chou2014). The TAM consists of three key variables: perceived ease of use (PEOU), perceived usefulness (PU) and potential users’ behavioural intention (BI) to adopt the technology in question (Chung et al., Reference Chung, Park, Wang, Fulk and McLaughlin2010). Meta-analysis results have indicated the robustness of paths from PEOU to PU and from PU to BI (Sun and Zhang, Reference Sun and Zhang2006; Chung et al., Reference Chung, Park, Wang, Fulk and McLaughlin2010). In addition, the TAM incorporates external and antecedent factors that affect the two key variables: PEOU and PU (Chung et al., Reference Chung, Park, Wang, Fulk and McLaughlin2010).
Hypothesis development
PU, PEOU, attitudes and reuse intention
According to the TAM, PEOU influences PU (Wang and Chou, Reference Wang and Chou2014) and attitude refers to the desirability to use IT (Shin, Reference Shin2010; Wang and Chou, Reference Wang and Chou2014). Improving the efficacy of e-business applications to enhance performance and efficiency positively affects users’ attitudes towards such applications (Aboelmaged, Reference Aboelmaged2010). Wang and Chou (Reference Wang and Chou2014) indicated that these factors enable users to perceive applications favourably because they are concerned with the effort required to use an application. Hence, this study proposed the following hypotheses:
• Hypothesis 1: PEOU positively affects PU.
• Hypothesis 2: PU positively affects attitude.
• Hypothesis 3: PEOU positively affects attitude.
Attitude refers to the desirability to use IT (Shin, Reference Shin2010; Wang and Chou, Reference Wang and Chou2014). Then, reuse intention refers to the likelihood of a person continuing to use an online system. This paper argues that when people realise the usefulness of an online system, they continue to use it. Kim and Park (Reference Kim and Park2005) illustrated that a consumer who favours a specific retailer is more willing to gather product information from that retailer. They confirmed that attitude positively affects search intention. In addition, people will continue use the e-business applications in order to help them control and collect information when they feel the e-business application is easy for them to use in the future. Accordingly, the present study proposed the following hypotheses:
• Hypothesis 4: PEOU positively affects reuse intention.
• Hypothesis 5: PU positively affects reuse intention.
• Hypothesis 6: Attitude positively affects reuse intention.
Health consciousness
Health consciousness assesses an individual's readiness to undertake health actions (Becker et al., Reference Becker, Maiman, Kirscht, Haefner and Drachman1977; Michaelidou and Hassan, Reference Michaelidou and Hassan2008). Health-conscious consumers are aware of and concerned about their wellbeing, and are motivated to improve or maintain their health and quality of life through engagement in healthy behaviours and consciousness of their health status (Gould, Reference Gould1988; Plank and Gould, Reference Plank and Gould1990; Kraft and Goodell, Reference Kraft and Goodell1993; Newsom et al., Reference Newsom, McFarland, Kaplan, Huguet and Zani2005; Michaelidou and Hassan, Reference Michaelidou and Hassan2008). Individuals tend to be aware of and interested in nutrition, physical fitness (Kraft and Goodell, Reference Kraft and Goodell1993; Michaelidou and Hassan, Reference Michaelidou and Hassan2008), food safety and other related information. In previous studies, individuals have identified health as a primary motivation to purchase organic food products (Grankvist and Biel, Reference Grankvist and Biel2001; Lockie et al., Reference Lockie, Lyons, Lawrence and Grice2004); studies have also found that health consciousness can predict attitudes and intentions to purchase organic foods because organic produce consumers are aware that food intake affects their health (Magnusson et al., Reference Magnusson, Arvola, Hursti, Åberg and Sjödén2003, Reference Magnusson, Arvola, Koivisto Hursti, Åberg and Sjödén2001; Michaelidou and Hassan, Reference Michaelidou and Hassan2008). Furthermore, Wang and Chou (Reference Wang and Chou2014) indicated that these factors enable users to perceive some applications favourably based on concern about the effort required to use an application, and favourably perceived applications further influence individuals’ attitudes. The present study argued that individuals with high health consciousness tend towards rational thinking, whereas those with low health consciousness tend towards emotional thinking. When individuals with high health consciousness perceive an e-business application as useful or easy to use, they feel that the product in question tends towards high quality and generates a positive product attitude. By contrast, when individuals with low health consciousness perceive an e-business application as not usefulness or not easy to use, they feel that the product in question tends towards low quality and generates a negative product attitude. Accordingly, this study proposed the following hypotheses:
• Hypothesis 7: Interactions between PU and health consciousness affect attitude; individuals with high health consciousness and high PU have more positive attitudes than those with low health consciousness and low PU.
• Hypothesis 8: Interactions between PEOU and health consciousness affect attitude; individuals with high health consciousness and high PEOU have more positive attitudes than those with low health consciousness and low PEOU.
Trust and attitude
In social psychology, trust refers to the belief that other people will react in predictable manners and the belief that one may rely on promises made by others (Pavlou, Reference Pavlou2003; Kim et al., Reference Kim, Kim and Shin2009). In e-commerce, eTrust refers to ‘a general belief in an online seller that results in behavioral intention’ (Gefen, Reference Gefen2000), combining trustworthiness, integrity and benevolence, all of which increase BI in inexperienced consumers through reduced risk (Jarvenpaa et al., Reference Jarvenpaa, Tractinsky and Saarinen1999; Kim et al., Reference Kim, Kim and Shin2009). Thus, trust positively affects attitude. Accordingly, this study proposed the following hypothesis:
• Hypothesis 9: Trust positively affects attitude.
Research method
Sampling and data collection
Figure 1 demonstrates the research framework. The present study conducted a paper survey because the elderly study participants all had experience of using traceability QR codes and a paper survey was more suitable for targeting such respondents. The paper questionnaires were distributed in a market survey from 15 March to 5 May 2016, and recruited volunteers to join this survey from senior citizens learning camp and community college in Taipei City. To control for product effects across the product categories of traceability QR codes, this study selected agricultural products because most traceability QR codes are connected with agricultural products. For target respondents with no prior experience of using traceability QR codes, this study provided codes and product images on the survey questionnaire. A total of 393 usable responses were obtained. The statistical results showed that 229 of the respondents were women (58.3%) and 164 were men (41.7%). The respondents were all aged 50–90 years and the average age was 57.5 years. Regarding purchase experience, 59.8 per cent of the respondents had experience of purchasing agricultural products by using traceability QR codes.
Research framework.
Measurements
This study conducted a questionnaire and based the measurement items on an extensive review of related studies to ensure content validity; items from the original scale were also used. PU, PEOU, attitude and reuse intention were measured using the four three-item scales in Wang and Chou (Reference Wang and Chou2014), and trust was measured using the seven-item scale in Gefen et al. (Reference Gefen, Karahanna and Straub2003). Health consciousness was measured using the six-item scale in Jayanti and Burns (Reference Jayanti and Burns1998). All items were measured using a seven-point Likert scale (1 = strongly disagree, 7 = strongly agree). All the items from the scales are presented in Table 1.
Internal reliability and convergent validity test results
Notes: CR: composite reliability. AVE: average variance extracted.
Data analysis and results
Before structural equation modelling (SEM) was performed, the samples were assumed to follow a multivariate normal distribution without multicollinearity. Kurtosis and skewness values were examined to ascertain whether the data were distributed normally and to avoid influencing the model estimation and test results. Hair et al. (Reference Hair, Anderson, Tatum and Black2006) proposed that a significant departure from normal distribution occurs when the corresponding skewness and kurtosis values fall outside the range of −1 to + 1. In this study, skewness was between −0.976 and −0.112 and kurtosis was between −0.600 and 1; both were within the acceptable ranges of the skewness and kurtosis indices, thereby signifying that the samples were normally distributed.
Confirmatory factor analysis
Confirmatory factor analysis (CFA) was conducted on all of the 45 items to determine whether the measurement variables accurately reflected the hypothesised latent variables (Table 2). An advantage of CFA is its ability to evaluate the construct validity of a proposed measurement theory (Hair et al., Reference Hair, Anderson, Tatum and Black2006). Construct validity determines whether measurement items are measuring the theoretical constructs that they claim to be measuring, and it includes convergent and discriminant validity. Hair et al. (Reference Hair, Anderson, Tatum and Black2006) indicated that convergent validity includes ‘individual item reliability’, ‘composite reliability’ and ‘average variance extracted’.
Correlations among latent variables
Notes: PU: perceived usefulness. PEOU: perceived ease of use. RI: reuse intention. HC: health consciousness.
Significance level: ** p < 0.01.
Model fit assessment and results
This study employed SEM to test the hypothesised model. The fit indices indicated that the hypothesised model fit the data well (Tsai et al., Reference Tsai, Cheng and Chen2011). The absolute-fit measure was adopted to assess how closely the theory in Hair et al. (Reference Hair, Anderson, Tatum and Black2006) fit the sample data. The chi-square to degrees of freedom ratio, goodness of fit index (GFI), adjusted GFI (AGFI) and root mean square error of approximation (RMSEA) were used to test the model. The results obtained using the conceptual model indicated that the model fit the data well, with chi-square (730.892) to degrees of freedom ratio = 142, GFI = 0.827, AGFI = 0.768 and RMSEA = 0.103. All fit measures and indices in this study exceeded the acceptable benchmarks, thereby indicating excellent model fit.
Table 3 shows the results of hypothesis testing. PU (p < 0.001), PEOU (p < 0.001) and attitude (p < 0.001) significantly affected reuse intention, indicating that Hypotheses 4–6 were supported. PU (p < 0.001), PEOU (p < 0.001) and trust (p < 0.001) significantly affected attitude, indicating that Hypotheses 2, 3 and 9 were supported.
Research findings
Notes: PU: perceived usefulness. PEOU: perceived ease of use. TR: trust. AT: attitude. RI: reuse intention. HC: health consciousness.
Significance level: *** p < 0.001.
Results for moderation of health consciousness
A two-way analysis of variance was performed on attitude to show the regression between PU and health consciousness (t(1, 392) = −3.871, p < 0.001). The main effects of PU (t(1, 392) = 13.877, p < 0.001) and health consciousness (t(1, 392) = 6.990, p < 0.001) were significant.
This study further used procedures developed by Aiken and West (Reference Aiken and West1991) and Dawson and Richter (Reference Dawson and Richter2006) to plot the interaction effects for significant differences between slopes, and found that the participants with high health consciousness and high PU had more positive product attitudes than those with low health consciousness and low PU, thereby indicating that Hypothesis 7 was supported (Figure 2).
Effect of the interaction between perceived usefulness (PU) and health consciousness (HC) on attitude.
This study performed a two-way analysis of variance on attitude to show the regression between PEOU and health consciousness (t(1, 392) = −2.951, p = 0.003 < 0.05). The main effects of PEOU (t(1, 392) = 14.200, p < 0.001) and health consciousness (t(1, 392) = 6.233, p < 0.001) were significant. The interaction effects obtained by Aiken and West (Reference Aiken and West1991) and Dawson and Richter (Reference Dawson and Richter2006) were plotted, revealing that the participants with high health consciousness and high PEOU had more positive attitudes than those with low health consciousness and low PEOU, thereby indicating that Hypothesis 8 was supported (Figure 3).
Effect of the interaction between perceived ease of use (PEOU) and health consciousness (HC) on attitude.
Discussion
The main objectives of this study were to develop an integrated extensibility model incorporating the TAM and to investigate the impact of health consciousness on elderly adults’ acceptance of technology in relation to traceability information websites in Taiwan. In addition, this study conducted SEM to analyse the data. Recently, because of Taiwan's implementation of a production and marketing traceability system for agricultural products, most producers use this system to display their product information, and consumers can use the system to seek information about agricultural products.
Because of the ageing population in Taiwan, concerns over health are on the rise; this leads to a demand to consume healthier food. Elderly people are increasingly paying close attention to health and developing high health consciousness. Thus, delivery of accurate health information is crucial. This study explored how health consciousness affects the attitudes of elderly people when they are using an agricultural product traceability system. The results revealed that elderly people with high health consciousness and high PU had more positive attitudes than those with low health consciousness and low PU, and those with high health consciousness and high PEOU had more positive attitudes than those with low health consciousness and low PEOU. In addition, the results of this study confirmed the effects of PEOU on PU; those of PU and PEOU on trust and attitude; and those of trust, PU, PEOU and attitude on reuse intention. All of these effects are considered in the TAM and have been mentioned in related studies (Jarvenpaa et al., Reference Jarvenpaa, Tractinsky and Saarinen1999; Kim et al., Reference Kim, Kim and Shin2009; Aboelmaged, Reference Aboelmaged2010; Wang and Chou, Reference Wang and Chou2014).
The results presented in this paper have academic and practical implications. Regarding academic implications, the major contributions of this research are the incorporation of health consciousness into the TAM and the exploration of attitude and reuse intention in the agricultural product traceability system for elderly people; neither of these aspects had been explored in previous studies. The present study hypothesised that health consciousness among users of a website usually affects consumer behaviours in real life; in particular, individuals’ health-related wellbeing and motivation affect their health improvement or maintenance. This study found that individuals who exhibited high health consciousness, high PU and high PEOU while using the agricultural product traceability system had more positive attitudes and higher reuse intention. This study also indicated that individuals are influenced by the PU of a website. Most related studies have investigated young or middle-aged consumers with experience of using the traceability system; by contrast, few studies have investigated the use of this system among elderly people. The present study focused on elderly people and analysed health consciousness among elderly people to verify that such people can increase their health consciousness through activities that encourage healthy ageing. Because internet use among older adults is gradually increasing and becoming more mainstream, older adults are becoming more likely to seek health information and purchase healthy products but less likely to watch videos, download music, play games and read blogs online (Jones and Fox, Reference Jones and Fox2009; Chung et al., Reference Chung, Park, Wang, Fulk and McLaughlin2010). Thus, this study incorporated heath consciousness into the TAM.
Regarding practical implications, the following suggestions are provided for policy makers and marketing practitioners. Health consciousness affects elderly people's attitudes towards using the traceability system. This study suggests that training courses or social activities should be offered to teach health information to elderly adults in order to promote health-related wellbeing and motivation, and also to teach them how to use technological applications and the traceability system to seek health information. Training courses could strengthen the learning behaviours of elderly adults’ partner groups, thereby increasing their degree of acceptance for using innovative technology and sharing knowledge, and further enhancing their health consciousness and ability to use IT. Health consciousness is a key topic, and marketing practitioners could attempt to employ experiential learning methods to promote health consciousness and social learning among elderly adults and further increase their usage and acceptance of new knowledge and life skills. Most elderly adults still purchase agricultural products at physical stores and markets. Elderly people are advised to pay attention to health information and food safety as they continue to age. The production and marketing traceability system can assist elderly consumers in understanding information regarding production processes and management of agricultural products, thereby increasing their trust in agricultural products through accessible and transparent information.
Limitations and future research
Although this study contributes to elderly consumer research, it had several limitations. This paper offers recommendations for future research based on these limitations. First, elderly consumers are generally unfamiliar with TAPs in Taiwan because TAP prices tend to be higher than those of general agricultural products. Thus, elderly consumers generally purchase cheaper general agricultural products or organic foods that they like. Therefore, such consumers seldom use traceability systems. Second, despite the samples in this study having been collected from elderly Taiwanese consumers – who have different degrees of health knowledge and health motivation – the collected sample data could not represent all elderly consumers. Thus, future studies should consider such differences in degrees of health knowledge and health motivation. Third, this study investigated behaviours of elderly adults while using the TAFT website. Future studies should analyse consumers from a wider age range to highlight age-related differences.
Acknowledgements
The authors thank the Ministry of Science and Technology, Taiwan for financial support.
Financial support
This work was supported by the Ministry of Science and Technology, Taiwan, ROC (grant number MOST 107-2410-H-003-034).
