*2.1. Short-Video Advertising*

Digital advertising significantly influences users' attitudes and intentions to purchase [5]. There are many studies about the factors that influence the effectiveness of online advertising [6], video advertising [7], and social media advertising [2]. Although short-video ads are a promising form of online advertising, this ad form has not been adequately examined in the research. The core of short-video ads is "video + sociality", and it has become an important field for the expansion of social media. Therefore, we study our questions mainly based on the research of social media advertising. Some quote video studies, of course.

There have been some existing studies on informativeness and intrusiveness and how such concerns may influence users' attitudes toward advertising and product purchase intentions [8–16]. Generally, informativeness is defined as the information value provided by media like texts, images, or videos [11] and intrusiveness is defined as the degree to which people deem the presentation of information as contrary to their goals [14]. Both informativeness and intrusiveness have been regarded as key factors to predict the helpfulness of messages in the research of social networks. For example, Dehghani et al. [12] conducted an empirical study on YouTube ads and found that a high value of informativeness led to a high level of user acceptance of YouTube ads, meaning that informativeness is a positive factor affecting users' attitude toward YouTube ads. Meanwhile, they found intrusiveness is a negative factor, meaning that a high value of intrusiveness led to a low level of user acceptance of YouTube ads. Lee and Hong empirically investigated that informativeness has a potential contribution to a positive online behavior based on the theory of reasoned action and the social influence theory [13]. Jung found perceived ad relevance influences advertising effectiveness while it could increase privacy concerns, which ultimately raise ad avoidance in social media [16].

Relevance is another factor that may affect user acceptance of advertising. Relevance has been widely adopted in web search and online recommendation [17]. For instance, a web search engine can rank the searching results according to the relevance between the user query and the returned web pages. In the online recommendation in E-commerce, the relevance between the user profile and products is usually used to predict the items that may fit within the users' interests. Similarly, the relevance between a user's interest and short-video advertisements may also affect the user acceptance of short-video ads [18]. Therefore, we will also consider relevance as a construct in the research model of this study, which will be detailed in Section 3.

On the other hand, according to the social influence theory [19], a high level of social interactions can enhance the feelings of a ffection, trust, belongingness, and warmth. Currently, most people are engaged in short videos by mobile apps. The mobile short-video apps o ffer people real-time interaction with other ones. In addition, platforms like Douyin and Tik Tok allow users to interact with short videos. For example, users can click the product link and fill a form to deliver their purchase interest to vendors. Therefore, social interaction can be a key factor a ffecting user acceptance of short-video ads.

So far, short-video ads are still in the early developing stage. Existing work has paid little attention to how the new features of short videos impact users' attitudes toward purchase behavior. Based on previous experiences in social media ads, many researchers claimed that two aspects of information should be considered in the context of social media ads [20–22]. The first one, product information, focuses on informativeness, relevance, and intrusiveness of information [20]. The second one, personal interaction, refers to social interaction, which reflects the impact of social identity and group norms on community users' group intention of accepting advertising [21]. Windels et al. examined the di fferences between native advertisements and friend referrals on social networking sites. They found that social relationships did not always work [22]. In this paper, we consider these related studies as well as the features of short-video ads to design antecedents toward the perceived usefulness of users and the attitude to short-video ads. As a result, four antecedents including informativeness, relevance, intrusiveness, and social interaction are designed in our study.

#### *2.2. Technology Adoption Model (TAM)*

The Technology Acceptance Model (TAM) was developed to describe users' behavior to accept or reject the use of new technologies [23]. The TAM model defines two variables, namely perceived usefulness and perceived ease of use, to quantify user attitude to information technology, which in turn can be used to measure user acceptance of information technologies.

Although the TAM model is initially designed to explain and predict the behavior of individuals on the use of information systems, it has been used in many studies [24–26]. Previous studies have demonstrated the applicability of the TAM model to online advertising. For example, Demangeot and Broderick examined a modified technology acceptance model (TAM) that was developed to test the intention to use SMS advertising [27]. Based on the ability of TAM and short video including a wide variety of online media services and SNS, it is a reasonable choice to consider the TAM model to model and quantify user acceptance of short-video ads.

The TAM model has been widely adopted in many existing studies. In this paper, we also adopt the TAM model as the basic research model. There are two reasons. First, many studies in social media have applied the TAM model to analyze the acceptance factors of online advertising and social network advertising. The research scope of this paper, i.e., short video, has the characteristics of both online media and social networks; thus, it is a reasonable choice to select the TAM model. Second, the basic factors in the TAM model, namely perceived usefulness (PU) and perceptual ease of use (PEOU), are suitable to distinguish the usefulness of live advertising and the impact of live-video technologies.

On the other hand, di ffering from the traditional TAM model, we integrate social interaction, intrusiveness, informativeness, and relevance with TAM and propose to use perceived usefulness as the mediate factor, forming an augmented TAM model that is more suitable for studying the intact path of short-video ads. Even though research integrating these external variables with the TAM is limited, available empirical findings generally support the influence of these factors on perceptions about users' ad attributes and adoption intentions.

In addition to the TAM model, another widely used model for predicting user acceptance of short-video ads is the Unified Theory of Acceptance and Use of Technology (UTAUT) model [28], which was proposed to incorporate various models of human behavior theory. The UTAUT model was constructed by extracting three variables that a ffect users' behavioral intentions, one variable

that influences action, and four moderators that mediate the e ffects of the process. Some of the variables had similar concepts with variables that construct the TAM model. One major di fference between UTAUT and TAM was that UTAUT proposed four control variables (i.e., gender, age, experience, and voluntariness) to further enhance the predictive power of the model. Based on UTAUT, Venkatesh et al. incorporated three other constructs into UTAUT, namely hedonic motivation, price value, and habit, extending UTAUT into UTAUT 2 [29]. Compared to UTAUT, the extensions proposed in UTAUT2 produced a substantial improvement in the variance explained in behavioral intention. As a result, both UTAUT and UTAUT 2 have been actively used in predicting users' purchase behavior in many areas like healthcare systems [30] and social-network-based advertising [1]. However, as suggested by Venkatesh et al. [29], in order to increase the applicability of the original UTAUT as well as the UTAUT 2 model, other relevant factors are usually needed to be extended to meet the specific requirements of di fferent applications, technologies, countries, etc.

In summary, both TAM and UTAUT are actively utilized in the research of user adoption of new technologies. Whether or not to use TAM or UTAUT mainly depends on the particular type of the application. In this paper, we select TAM rather than UTAUT as the basic model for analyzing user acceptance of health-related short-video ads for the following reasons. First, UTAUT is a behavioral model aiming to explain the behavior of people in the use of information systems [31], but in this study, the content of short-videos, as measured by the antecedents named informativeness, relevance, and intrusiveness in our model, is also an important factor. Second, the UTAUT model is more complex than TAM; thus, extending TAM is easier than extending UTAUT. Nevertheless, as Venkatesh et al. [29] and Dwivedi et al. [32] reported, most studies only employed a subset of the UTAUT model, and the moderators were typically removed. To this end, the TAM model can have similar predictive power as the UTAUT model without moderators. However, it is still worth investigating the extension of the UTAUT or UTAUT 2 model to predict the user adoption of health-related short-video ads in the future.

#### **3. Research Model and Hypotheses**

### *3.1. Research Model*

In this paper, we first selected the main factors in the TAM model as basic independent variables, i.e., perceived usefulness (PU), perceived ease of use (PE), and ad attitudes (AT). Further, we introduced social interaction (SI), informativeness (IR), intrusiveness (IN), and relevance (RE) as new antecedents of perceived usefulness. The selection of the four factors (i.e., SI, IR, IN, RE) was mostly based on two aspects. First, previous studies in social media have shown that these factors are critical to the usefulness of social media advertising [2,12,16]. Second, these factors are suitable for short-video ads, as they reflect the information features of live video ads. Thus, it is reasonable to introduce them to the research model. As a result, the perceived usefulness (PU) in our research model became a mediator variable. The purchase intention (PI) was set to be the dependent variable. In the study of behavior intention, attitude has been commonly recognized as a variable that a ffects the willingness of behavior. Thus, we chose attitude as another mediator variable. On this basis, this paper put forward the research model of user responses to health-related short-video advertisements. Figure 2 shows the components of the proposed research model.
