Measures of Perceived Interactivity: An Exploration of the Role of Direction of Communication, User Control, and Time in Shaping Perceptions of Interactivity
By Sally J. McMillan and Jang-Sun Hwang
University of Tennessee
Revised and Resubmitted to the Journal of Advertising
For Special Issue on Advertising and The New Media
Contact information for the first author:
476 Communication Building
University of Tennessee
Knoxville, TN 37996-0343
Phone: 865-974-5097
Fax: 865-974-2826
E-mail:
This study was funded in part by a grant from the University of Tennessee Scholarly Activities Research Initiative Fund (SARIF).
Measures of Perceived Interactivity: An Exploration of the Role of Direction of Communication, User Control, and Time in Shaping Perceptions of Interactivity
Abstract
Interactivity is an often-mentioned but seldom-operationalized concept associated with the World Wide Web. Interactivity has been positioned conceptually as a process, a function, and a perception but most operational definitions have focused on process or function. This study developed scales to operationalize the perception-based approach to interactivity because consumer perceptions are central to advertising research. Three overlapping constructs that are central to interactivity are explored: direction of communication, user control, and time. A multi-stage method was used for identifying and refining measures of perceived interactivity (MPI). The 18 items included in the MPI offer researchers a tool for measuring a consumer perception that is central to advertising on the Web.
The word “interactive” is often used as a synonym for new media such as the World Wide Web. Advertising practitioners and researchers use the phrase “interactive advertising” to describe Internet or Web-based advertising. However, despite widespread use of such terms, scholars have noted that interactivity is often either undefined or under-defined (Hanssen, Jankowski, and Etienne 1996; Heeter 1989, 2000; Huhtamo 1999; Miller, Katovich, and Saxton 1997; Schultz 2000; Smethers 1998).
A better understanding of interactivity is of critical importance for those who want to analyze and/or develop Web-based advertising. Bezjian-Avery, Calder, and Iacobucci (1998) indicated that new media fundamentally change relationships between consumers and producers by opening up the potential for new forms of dialogue. However, their study also found that (p.30): “the consumer can still simply buzz right through the interactive media, paying so little attention to the advertisements, that the message cannot function persuasively.” If advertising is to be persuasive in environments such as the Web, advertisers need to better understand interactivity so that they can more effectively engage and interact with consumers.
A central focus of advertising research is consumer perception. Researchers have begun to develop perceptual measures for the Web such as attitude toward the Web site (Chen and Wells 1999). But, to date, research on perceived interactivity of the Web has been primarily conceptual rather than operational or has used measures that have not been fully tested and sometimes have low reliability (Cho and Leckenby 1999; Heeter 2000; Lee 2000; McMillan 2000a; Sundar et al. 1998; Wu 1999). The purpose of this study is to operationalize measures of perceived interactivity in the context of the World Wide Web using the rigorous scale-development process applied to many marketing scales. Scales for perceived interactivity are needed to help researchers understand the role of interactivity and to help advertisers benefit from the potential of Web-based interactivity.
Literature
For the past 20 years, interactivity has been widely discussed in fields such as advertising, marketing, communication, information science, computer science, and education. Table 1 summarizes some key definitions of interactivity with emphasis on the advertising, marketing, and communication literature. Several examples of pre-Web definitions of interactivity are also included.
------
Table 1 About Here
------
Definitions of interactivity can be categorized based on the primary focus of authors on process, features, perception, or combined approaches. From the process perspective, scholars focus on activities such as interchange and responsiveness that are key to interactivity. Scholars who focus on features seek to identify either general characteristics (such as user control and two-way communication) or specific characteristics of Web sites (such as search engines and chat rooms) that define interactivity. Lee (2000) suggested that interactivity should not be measured by analyzing processes or by counting features. Rather, researchers should investigate how users perceive and/or experience interactivity. The focus on perception is consistent with marketing, advertising, and communication traditions. As Reeves and Nass (1996, p.253) noted: “Perceptions are far more influential than reality defined more objectively.” Little work has been done to operationalize perceived interactivity. This study seeks to fill that gap.
As illustrated in Table 1, interactivity has been defined based on multiple processes, functions, and perceptions. However, three elements appear frequently in the interactivity literature: direction of communication, user control, and time. These elements hold promise for exploration of perceived interactivity on the Web because they serve as umbrellas for many of the elements identified in Table 1. For example, direction of communication encompasses the concepts of responsiveness and exchange; user control includes functions such as participation as well as features such as search engines; and the concept of time embraces issues such as timely feedback and time required for information retrieval.
Direction of Communication
Researchers who examine ways that computers facilitate human interaction often focus on the importance of enabling two-way communication (Beniger 1987; Bretz 1983; Chesebro 1985; Duncan Jr. 1989; Durlak 1987; Garramone, Harris, and Anderson 1986; Kirsh 1997; Rafaeli and Sudweeks 1997; Zack 1993). Within some of the literature, two-way communication is characterized as mutual discourse (Ball-Rokeach and Reardon 1988; Burgoon et al. 2000; Hanssen, Jankowski, and Etienne 1996; Williams, Rice, and Rogers 1988). Other scholars focus on the capability for providing feedback (Day 1998; Duncan and Moriarty 1998; Ha and James 1998; Newhagen, Cordes, and Levy 1996).
Massey and Levy (1999) note that the Web provides for “interpersonal interactivity” because individuals can communicate with each other through tools such as chat rooms and bulletin boards. By making their Web sites friendly to users, marketers can facilitate this kind of interpersonal interactivity and can also generate positive word of mouth for their companies (Hoffman and Novak 2000). Interpersonal two-way communication through the Web also facilitates other kinds of business relationships such as customer service and supply chain management (Berthon, Holbrook, and Hulbert 2000; Peltier, Schibrowsky, and Davis 1998).
User Control
The Web often provides users with more content and more navigational tools than do traditional media. Much of the literature that focuses on human-to-computer interaction (HCI) examines the ways that humans control computers and other new media such as DVD and video games (Burgoon et al. 2000; Hanssen, Jankowski, and Etienne 1996; Huhtamo 1999; Milheim 1996; Murray 1997; Preece 1993; Tan and Nguyen 1993; Trevino and Webster 1992).
Reeves and Nass (2000) noted that some HCI studies focus on human perception and others on computer design. Studies with a human focus examine how individuals interpret computer personality (Moon and Nass 1996), the level of agency that individuals perceive they have in working with computers (Huhtamo 1999; Murray 1997), individual decision styles (Vasarhelyi 1977), and the goals that individuals bring to the system (Belkin, Marchetti, and Cool 1993; Xie 2000). Computer-focused studies examine issues such as interfaces and input devices (Baecker 1980; Biocca 1993; Laurel 1990; Naimark 1990; Nielsen 2000; Schneiderman 1998), navigation tools (Heeter 2000; Nielsen 2000), features for user choice and input (Belkin, Marchetti, and Cool 1993; Daft, Lengel, and Trevino 1987; Durlak 1987; Hanssen, Jankowski, and Etienne 1996; Looms 1993; Mahood, Kalyanaraman, and Sundar 2000; Steuer 1992; Zeltzer 1992), and system activity (Milheim 1996; Valacich et al. 1993).
Time
Perception of Web-based interactivity is influenced by the speed at which messages can be delivered and the speed at which individuals can process messages. Crawford (1990) noted that for interactive systems (p.105): “The ideal is to have the computer moving at a speed that doesn’t inhibit the user.” Speed of response is a central concern of both developers and users of interactive media (Dellaert and Kahn 1999; Kay 1990; Nielsen 2000; Vora 1998).
Another time element important to interactivity is the ability of users to quickly navigate through a wealth of information and easily find what they are seeking (Mahood, Kalyanaraman, and Sundar 2000; Nielsen 2000; Wu 1999). One study (Latchem, Williamson, and Henderson-Lancett 1993) noted a benefit of interactive systems is that users (p.23): “can work in their own time and at their own pace, choose their preferred navigational pathways and delivery systems and develop their own mental models and schemata.”
Overlapping Dimensions
Each of the three dimensions identified above is central to the concept of interactivity, but in much of the literature these concepts overlap and are interrelated. Communication and control overlap as higher levels of control lead to more active participation in communication. As Naimark (1990) noted, interactivity is often defined at the intersection of these two concepts (p.455): “Though interactivity always requires information flowing in both directions, it is our input and its effect that distinguishes it from non-interactivity.”
The intersection of time and communication is often viewed in the framework of whether interactive communication occurs in real time. Some have suggested that real-time, synchronous communication is central to the concept of interactivity (Kiousis 1999; Murray 1997; Steuer 1992; Straubhaar and LaRose 1996). Others have suggested that asynchronous communication, characterized by tools such as e-mail and newsgroups, is a key benefit of interactivity (Rheingold 1993; Williams, Rice, and Rogers 1988). Some studies have compared synchronous and asynchronous communication in the context of interactivity (Hesse, Werner, and Altman 1988; McGrath 1990; McMillan and Downes 2000; Morris and Ogan 1996; Walther 1992).
Time and control also overlap as complexity of control systems (e.g. navigation tools) affects the time required to access content. But time spent controlling the system is not always related to complexity – increased time can also be a result of intense engagement. Literature on the concept of “flow” (Csikszentmihalyi 1975; Ghani and Deshpande 1994; Hoffman and Novak 1996; Novak, Hoffman, and Yung 2000; Trevino and Webster 1992) focuses on how users can become absorbed in new media and lose track of time. Ghani and Deshpande (1994) defined flow as the activities that result from both the exploratory, and often time-consuming, use of the computer and the individual’s sense of being in control of the computer.
Figure 1 illustrates the overlapping of the three dimensions of perceived interactivity that are central to this study. Web-based interactivity involves communication among individuals, the ability those individuals have to control information and to participate in active communication, and time – time to load the message, time to find information, time to communicate with others, and the loss of time as one gets caught in the flow of computer-mediated communication. Users’ perceptions of direction of communication, control, and time are central to how interactive they perceive Web sites to be.
------
Figure 1 About Here
------
Scale Development
Churchill’s (1979) paradigm for scale development served as the primary guideline for developing a scale for measuring perceived interactivity. Other updated guidelines (e.g., DeVellis 1991; Gerbing and Anderson 1988) were also considered. A multi-method approach was used for specifying the domain of interactivity and generating an initial pool of items designed to measure perceived interactivity. Additionally, two pre-tests as well as two data-collection stages were used to generate and refine a set of scale items. This multi-stage process ensured construct validity by enabling the researchers to identify, purify, and assess potential scale items based on multiple types of qualitative and quantitative feedback.
Item Generation
Three sources were used to specify the domain of interactivity and to generate the initial item pool. First, the literature on interactivity suggested a conceptual model that identified three overlapping constructs that underlie interactivity (see Figure 1). Second, in-depth interviews with 10 individuals who produce and teach interactive communication were used to probe for understanding of the concept of interactivity (McMillan and Downes 2000). Transcripts of those interviews were reviewed and all words and phrases used in conjunction with interactivity were identified. Finally, two focus groups were conducted. In these focus groups, individuals identified and discussed words and concepts that they associated with interactivity.
From the literature, expert interviews, and focus groups, 157 unique words and phrases emerged. Upon examination, some words and phrases were eliminated because they addressed concepts related to, but not central to, interactivity (e.g. words for specific technologies such as computers and digital TV). Other terms were closely related and could be summarized with two to three synonyms rather than the 10-15 different terms on the original list. Additionally, the researchers identified antonyms for each core concept. The purpose of these antonyms was to ensure that study participants were actually reading all items rather than simply selecting one rating and applying it to all items. Correlations between the antonyms and related terms in each step of the analysis yielded strong negative correlations. This indicates that respondents did differentiate antonyms from words positively associated with interactivity.
The list of 157 words and phrases was reduced to 26 core concepts. Five core concepts were associated with Communication, five with Control, and six with Time. Ten additional core concepts addressed areas of overlap between dimensions. While some concepts in the original list were eliminated, these 26 core concepts were most central to the literature and the experts and were also frequently mentioned in focus groups.
Data Collection
Based on a pre-test with 60 undergraduate students, computers were chosen as the topic for exploration of interactivity because they scored well on three criteria: likelihood of seeking information, likelihood of communicating with others about the topic, and likelihood of online purchase. Two Web sites were developed to provide an environment in which to evaluate perceptions of interactivity. Laptop computers were the subject of both Web sites; the sites contained virtually identical information. However, one site was designed to have fewer interactive features and fewer opportunities for interactive exchange thus making it likely to generate lower scores on a scale of perceived interactivity than the other site that included features such as chat rooms, bulletin boards, a site map, and enhanced navigation bars. The purpose for creating these two environments was to generate variance in response to measures.