Keywords

1 Introduction

Decreasing fertility rates and extending life expectancy at birth have made aging population a serious issue in the world. Compared with that in 2017, the number of persons aged 60 or above is expected to more than double by 2050 and to more than triple by 2100, that is, rising from 962 million globally in 2017 to 2.1 billion in 2050 and 3.1 billion in 2100 [1]. Population aging is projected to have a profound effect on societies where many countries are likely to face increasing fiscal and political pressures from health care, old-age pension, and social protection systems [1].

Gerontechnology aims to apply technology to assist in dealing with problems and difficulties arising from aging to give older people the chance to lead healthy, independent, and socially engaging lives on a continual basis [2]. Many previous studies corroborated that the use of gerontechnology by seniors has the potential to alleviate aging-related problems [3,4,5,6,7,8]. However, older people do not show as considerable enthusiasm for adopting new technologies as young people [9, 10]. For instance, they do not use smart phones and tablets as frequently as young people do. In mainland China, elderly people aged 60 or above accounted for only 4.8% of Chinese Internet users compared with 42.9% of adults aged 30 to 59 and 52.2% of young adults aged 29 or below [11]. Thus, the low adoption rate of technology limited the benefits of technology for older people. In addition, as technology becomes highly integrated in everyday life, not being able to use technology may put older adults at a disadvantage in terms of their ability to live and function independently and successfully perform daily tasks [12].

Acceptability is the key factor to integrate new technologies at home, particularly when the users are elderly or low information communication technology (ICT)-educated persons [13]. Technology acceptance means “the approval, favorable reception and ongoing use of newly introduced device and system” [14]. Fishbein and Ajzen [15] verified that a person’s behavior is driven by his or her intention to perform the behavior. Hence, behavioral intention is an immediate antecedent of behavior. Self-efficacy is a perceived capability to perform or learn behaviors at specific levels [16]. The achievement, self-regulation, and motivation of individuals can be powerfully influenced by self-efficacy [16]. People with high self-efficacy would have substantial interest and motivation in performing a certain behavior. Information technology acceptance is a behavior of an ongoing use of a newly introduced information device or system; hence, this behavior is closely correlated with behavioral intention and self-efficacy of using information technology.

A study of gerontechnology acceptance was done by Chen and Chan [17] with 1,012 seniors aged 55 or above in Hong Kong. They validated that individual attributes, such as age, gender, education, gerontechnology, self-efficacy and anxiety, health and ability characteristics, and facilitating conditions, are explicit and direct predictors of technology acceptance. Chen and Chan [17] also confirmed the importance of environmental supports (e.g., financial support and training opportunities) in helping Hong Kong elders overcome barriers to technology. A study of acceptance of smart phone technology by older Chinese adults was done by Ma et al. [18] with 120 Chinese older adults aged 55 or above. Evidently, facilitating conditions are important factors influencing the perceived ease of use, and the opportunity of getting external help is highly important for smartphone acceptance. Few other studies stressed the importance of attending training programs or workshops for older individuals to build self-confidence, elicit positive attitudes, and increase the intention of using technology [19, 20].

Social cognitive theory (SCT) has been widely used in understanding the process of knowledge acquisition and information technology training in psychology and education studies [21,22,23]. This theory states that people acquire knowledge from observing others performing a behavior and its consequences. The subjects to be observed and imitated are termed role models. Gupta et al. [24] confirmed that observational learning leads to better performance compared with traditional lecture with passive learning. In addition, observational learning is associated with positive reported evaluation, less negative effect, and great satisfaction during training. Conducting observational learning through a video-taped demonstration with a human model is popular and cost-effective [22, 25]. Research on technology training for older adults have investigated the self-reported experiences of older people and their training preferences [26, 27], but how observational training influence older adults’ information technology acceptance has not been studied yet.

Given that training environments are critical for examining expected training outcomes [27], the impact of human elements, such as models, instructors, and learning peers, are worth considering. Studies affirm that older adults experience difficulty when receiving technical support from young people and that they profoundly learn when helped by their peers. This scenario is probable when their peers have similar problems and considerable understanding of the difficulties faced by older adults when using technology [28, 29]. However, certain studies recommend facilitations from younger generations, such as children and grandchildren. For instance, Fausset et al. [30] suggested letting the older adults’ children play a major role in influencing technology adoption and use. Lin et al. [31] corroborated that older adults chose younger children and friends with skills they admired as their role models in learning to use the Internet. According to the aforementioned statements, an investigation of the effects of model generation on facilitating the older adults’ information technology acceptance is needed.

In this study, behavioral intention and self-efficacy were taken as the measurements of information technology acceptance. Model generation was set as a factor that may influence training outcomes. The current study aimed to investigate the impact of observational learning with live models on enhancing older adults’ information technology acceptance and to study the effects of model generation on training outcomes.

2 Methodology

2.1 Experimental Design and Hypotheses

An experiment of observational learning with a video-taped demonstration in between-groups design was developed. Three training groups (child, young adult, and old adult) were included in this experiment. Model generations consisted of the child model with demonstrators aged 6–14, the young adult model with demonstrators aged 18–25, and the old adult model with demonstrators aged 60–70. One female and one male demonstrators were involved in each generation model group. Sixty participants were randomly assigned to the three training groups with twenty participants in each generation group. Other than different model generations, each participant in the three groups received the same training content and procedure.

The following hypotheses were tested in this experiment:

H1: :

Observational training with a video-taped demonstration is effective in improving the self-efficacy of older adults toward using information technology.

H2: :

Observational training with a video-taped demonstration is effective in improving the behavioral intention of older adults toward using information technology.

H3: :

Observational training outcomes, including self-efficacy and behavioral intention, are significantly influenced by model generation

2.2 Participants

A total of 60 participants, including 33 females and 27 males, were recruited. All were aged 60–80 (M = 68.4, SD = 5.17) without severe cognitive impairments. Cognitive function was measured by the Mini-Mental State Examination (MMSE), and older participants with scores of 18 or above were involved in the experiment, as suggested by Tombaugh and McIntyre [32]. The participants have little (less than 1 week) or no experience in using tablets and apps. They signed an informed consent before joining the study.

2.3 Apparatus and Training Materials

The apparatus used in the experiment included a notebook computer, tablets (Samsung Galaxy Note 10.1 N8000, size of 10.1 in. with 62.7% screen-to-body ratio), and pre-observational and post-observational learning questionnaires. The participants took observational learning by watching a video on the notebook computer. The video was about the daily use of three different apps [YouTube (video-sharing app), WhatsApp (social app), and Mass Transit Railway (mobile app)]. The tablets were used as platforms for information technology operation. The videos were recorded by three types of generation models. All models had at least one year of experience using tablets and the three apps to maintain homogeneity. In addition, models were required to perform actions by following a given script to ensure that the video content was standardized. Questionnaires were used to measure the participants’ self-efficacy and behavioral intention before and after the observational learning. Eleven and five statements using a 7-point Likert scale were included in the self-efficacy and behavioral intention assessment part. A 7 point refers to “extremely agree” with the statement, and a 0 point refers to “extremely disagree” with the statement. For instance, one of the statements for self-efficacy measurement was “I am confident in using a tablet by myself.” If the participants extremely agreed with the statement, they could give a score of 7. However, if they extremely disagreed with the statement, they could give a score of 0. The questionnaires for pre-observational learning and post-observational learning were the same.

2.4 Procedures

Before the observational learning, the participants were asked to fill the pre-observational learning questionnaire. Thereafter, they were required to learn how to use the three apps by watching the video displayed on the notebook computer. After watching, they imitated the steps without the video showing. After they finished their observational learning, they were requested to complete the post-training questionnaire. The testing procedure took approximately 20 min.

2.5 Data Analysis

Factor analyses, evaluations of descriptive statistics, and analyses of variance (ANOVA) for scores of self-efficacy and behavioral intention were performed with the use of the SPSS 22.0 software.

3 Results

3.1 Reliability and Validity of Measurements

Before data analyses, the reliability and validity of measurements for self-efficacy and behavioral intention were evaluated. The Kaiser-Meyer-Olkin measure of sampling adequacy (MSA) was 0.65, which was higher than the criterion value of 0.5 for being acceptable for factor-analytic purpose [33]. The communalities were all above 0.3, and the Bartlett test was significant. These scenarios verified that the sampling satisfied the conducting of factor analysis. Principal component analysis was used to calculate factor loadings of the retained factors (self-efficacy and behavioral intention), and the results showed that they were higher than the satisfied criteria value of 0.5. The composite reliability was examined to test the internal consistency and reliability of measurements. The values of the two constructs were higher than a value criterion of 0.6 [34]. All the results satisfied the requirements of convergent validity and reliability.

3.2 Demographic Information

Demographic data, including gender, age, marital status, education, working arrangement, self-assessed economic status, living arrangement, and tablet experience, were collected from the participants. Majority were married (68.3%), received primary education (31.7%), retired (88.3%), with general economic status (75%), and lived with household members (68.3%). Table 1 exhibits the detailed demographic information.

Table 1. Demographic information and tablet use experience (N = 60)
Full size table

3.3 Effectiveness of Observational Learning

Effectiveness of Observational Learning for the Entire Sample

The effectiveness of observational learning on facilitating technology adoption was measured with a paired t-test by comparing self-efficacy and behavioral intention before and after the training. Significant enhancements for self-efficacy (t(59) = −6.69, p < 0.05) and behavioral intention (t(59) = −4.39, p < 0.05) emerged after the training. The average score of self-efficacy after the training was 19.1% higher than that during the pre-training. In addition, the average score of behavioral intention after learning was 9.9% higher than that in pre-learning. Figure 1 illustrates the differences of self-efficacy and behavioral intention between pre-training and post-training.

Fig. 1.
figure 1

Differences of self-efficacy and behavioral intention between pre-learning and post-learning

Full size image

Effectiveness of Observational Learning for Three Generation Model Groups

All the three model groups of different generations significantly increased in self-efficacy after learning (p < 0.05). Specifically, the self-efficacy in post-training was 16.4%, 15.7%, and 26.1% higher than that in pre-training for the children, young adult, and old adult model groups (Fig. 2). Regarding behavioral intention, a significant increase of 14.6% occurred in the children model group (p < 0.05). No significant differences for the young adult and old adult model groups emerged between pre-training and post-training (Fig. 3) with corresponding increase rates of 7.0% and 2.5%, respectively.

Fig. 2.
figure 2

Differences of self-efficacy between pre-learning and post-learning for three generation model groups

Full size image
Fig. 3.
figure 3

Differences of behavioral intention between pre-learning and post-learning for three generation model groups

Full size image

3.4 Effects of Model Generation

Before the observational training, the sixty participants were randomly divided into three generation model groups, with no significant differences in self-efficacy (F(2, 57) = 2.38, p = .102) and behavioral intention (F(2, 57) = 1.55, p = .221). The effects of model generation on training outcomes were tested with ANOVA by comparing the mean values of self-efficacy and behavioral intention in post-training among the three generation model groups. The results confirmed the absence of statistically significant differences in self-efficacy (F(2,57) = 2.18, p = .122) or behavioral intention in post-training (F(2, 57) = 2.62, p = .081) among the groups.

3.5 Moderating Role of Demographics

The moderating effects of the participant demographics (gender, age, tablet experience, and living arrangement) on self-efficacy and behavioral intention after the training were tested. The effects of gender were tested with the t-test by comparing the mean values of self-efficacy and behavioral intention in post-training between the male and female participants. Linear regression was done for self-efficacy and age to detect whether self-efficacy linear dependence on age occurred. Similarly, linear regression was done for behavioral intention and age. The moderating effects of tablet experience and living arrangement on training outcomes were tested with the t-test and ANOVA. The results indicated that all the demographics of participants in gender, age, tablet experience and living arrangement had non-significant effects on self-efficacy and behavioral intention after the training.

4 Discussions

Technology training has been recommended as a facilitating condition to improve technology adoption for older people [17, 18]. Observational training with a video-taped demonstration is popular and cost-effective. In this study, the self-efficacy of the entire sample significantly improved after the training. In addition, a statistically significant improvement of self-efficacy occurred in all the three generation model groups. The result supported the research [25, 35] that observational learning with modeling provides a method of strengthening self-efficacy. H1 was supported.

Given that significant differences in behavioral intention between pre-training and post-training for the entire sample were observed, H2 was supported. Evidently, older adults had a higher intention of adopting information technology in post-learning than that in pre-training. A significant increase in behavioral intention was observed in the child model group (p < 0.05), but no significant increases in behavioral intention for the young adult or the old adult model group were observed. The results supported the statement of Luijkx et al. [36], which confirmed that allowing grandchildren to educate older adults on using information technology is worthwhile because older adults easily adopt their enthusiasm and may be persuaded to use technology.

No significant differences in self-efficacy or behavioral intention of post-training among the three generation model groups were observed. H3 was not supported. However, differences in increase rate of self-efficacy and behavioral intention among the three groups were evident. The greatest improvement of self-efficacy occurred in the old adult model group, which indicated that they had the best performance on improving self-efficacy for old technology learners. Schunk [37] corroborated that success of instructors with similar characteristics can raise the observers’ self-efficacy because they believe that they have a similar ability to be as successful as the performers. In the elderly model group, the video demonstrators and the learners were of similar age, and their successful operation of technology devices can highly enhance the learners’ confidence. Although the old adult model group made the greatest self-efficacy improvement with observation learning, behavioral intention did not consistently improve. The greatest improvement of behavioral intention was found in the child model group. According to planned behavior theory, human action is guided by the following three kinds of consideration: (1) beliefs about the likely outcomes of the behavior and evaluations of these outcomes (behavioral beliefs), (2) beliefs about the normative expectations of others and motivation to comply with these expectations (normative beliefs), and (3) beliefs about the presence of factors that may facilitate or impede the performance of the behavior and perceived power of these factors (control beliefs) [38]. Model generation may have an impact on normative belief to influence the behavioral intention of older learners. Children generally have high enthusiasm for and interest in information technology and their passions may impose high normative expectations on learners. Furthermore, older people tend to be motivated by children the most as child demonstrators are just like their grandsons/granddaughters. Thus, older learners had a significant improvement in behavioral intention with a child model.

The results of this study provide certain guidelines in technology training for older adults: (1) observational training with a video-taped demonstration is an effective method to improve technology adoption by older people, (2) old demonstrators can make the greatest improvement in an aging learner’s self-efficacy toward information technology use, and (3) child demonstrators can make the greatest improvement in an aging learner’s behavioral intention toward information technology use. In addition, the following limitations of this study should be considered: (1) the participants were recruited mainly from the elderly care centers of Tung Chung, Hong Kong due to limited resources and time efficiency consideration, which may have resulted in some selection bias of the sample, and (2) a tablet was chosen as the only platform to operate apps during the training because this devise is widely used and suitable for older people with vision impairments. However, mobile phones may be the most widely used devices for older people in practice. Other devices can be tested in the future.

5 Conclusions

The effectiveness of observational learning with live models in enhancing older people’s technology adoption was tested in this study. Self-efficacy and behavioral intention were the measurements of information technology acceptance. With observational training, the self-efficacy and behavioral intention of older people significantly improved. Statistically significant differences of self-efficacy between pre-training and post-training were observed in all the three generation model groups, but significant differences of behavioral intention were only found in the child model group. Observational learning with the old adult model had the best performance on improving self-efficacy for aging learners. However, observational learning with the old adult model produced the least effect on improving older people’s behavioral intention, where the greatest effect was found in the child model group. Demographic variables, including gender, age, tablet experience, and living arrangement of the participants, had no significant moderating effects on training outcomes. In conclusion, observational training with live models can increase technology adoption for older people, and a combination of child and old adult models is recommended.