THE FLORIDA STATE UNIVERSITY
COLLEGE OF COMMUNICATION
Beyond Demographics:Motivations of World-Wide-Web Users and
Their Implications for the Future of Advertising
By
Robert Bill Riley, Jr.
copyright 1997, Robert Bill Riley, Jr.
all rights reserved
A Research Proposal submitted to the
Department of Communication
in partial fulfillment of the
requirements for the degree of
Master's of Science
Degree to be Awarded:
Fall Semester, 1997
The members of the College of Communication Committee who have reviewed
the research proposal of Robert
Bill Riley, Jr. submitted on February 13, 1996.
C. Edward Wotring
Professor Directing Thesis
Edward J. Forrest
Committee Member
Daniel J. Montgomery
Committee Member
Jay D. Rayburn
Director, Graduate Studies
Barry S. Sapolsky
Chair, Department of Communication
John K. Mayo
Dean, College of Communication
,
Beyond Demographics:Motivations of World-Wide-Web Users and
Their Implications for the Future of Advertising
Name:Robert B. Riley, Jr.
Department:Communication
Major Professor:C.Edward Wotring
Degree:Master's of Science
Term Degree Awarded: Fall Semester, 1997
By 1990, only 67% of viewers were watching network TV. Each new technological advance has further fragmented broadcast consumption. New communication technologies such as the World-Wide-Web now allow the consumer to control exposure to advertising and marketing stimuli. This ability to control exposure severely reduces an advertiser's potential to reach large audiences and alters client perceptions about the value of advertising.
The dearth of reliable information about Web users hinders efforts to understand the advertising and marketing potential of this new technology on the broadcast industry. Little is known of how the informationally-empowered consumer chooses to interact with the Web or how this interaction influences responses to advertising and marketing communication. Understanding the determinants of consumer attention is the key to understanding the new media and therefore the focus of this study.
Mass marketers understand that exploratory behaviors such as information search is a key component of product purchase deliberation. The purpose of this study is to examine the relationship between exploratory behavior theory and information consumption behavior patterns by Web users. This study is important because it seeks to provide a theoretical explanation for Web usage when most current research has focused almost exclusively on discovering demographic information about Web users.
This study will use innovative new methods to examine the Web user population.
A Web site will be constructed to survey Web users in their natural environment.
This special audience research gathering technique has been proven to be
well suited to meet the new technology challenge by several pioneering
researchers. Any findings could be extremely useful to Web content providers
who are attempting to achieve a targeted communication goal.
The World Wide Web is an on-demand, consumer-oriented global information sharing network. It is the first example of a computer-mediated hypertext and multimedia environment replete with relatively standard software, protocols, and conventions which work across all computer platforms and configurations. The Web's unique capability is that it allows Internet users to search, retrieve, browse, and add graphical information at will (Hoffman & Novak 1995).
There are currently about 24 million Internet users in the United States and Canada. The Web is the most popular Internet resource (Nielsen Media Research, 1995) and gaining 150,000 new users each month (New York Times, 1993). Despite this extraordinary explosion of the Web user population, little is known about why people use the Web or what they feel about its tools and technologies. The dearth of reliable information about Web users confounds advertisers and marketers who attempt to achieve specific communication goals.
The purpose of this study is to examine the relationship between exploratory behavior theory and information consumption behavior patterns by Web users. This study is important because it seeks to provide a theoretical explanation for Web usage which moves beyond demographic profiles which are typical of Web research.
Little is known of how the informationally-empowered consumer chooses to interact with media or how this interaction influences responses to messages. Understanding the determinants of consumer attention is the key to understanding the Web's marketing and advertising potential and therefore the focus of this study.
The Internet is a distributed computer network which supports many-to-many communication on a global basis. The World Wide Web was conceived and developed for the Internet at the European Laboratory for Particle Physics (CERN). First released to the world in 1991, the Web represents "the universe of network-accessible information, an embodiment of human knowledge" (Berners-Lee etal.,1992).
Consumer oriented, Web users navigate through Web "sites" via hypertext links to server computers throughout the world. A server is simply a program that answers requests for documents from Web users over the Internet. Hypertext is a non-linear organization of information consistent with the Web's conceptual goal of providing universal readership.
The advent of the Web is widely accepted as the reason for the explosion in Internet traffic during the 1990's. The success of the Web is due to the ease in which even novice computer users can connect to, use, and add to the Web's global resources.
Among theories describing human behavior, the idea that people act merely out of desire to achieve a preferred level of stimulation has been widely investigated by researchers from both psychological and consumer behavior perspectives.
Optimum stimulation level (OSL) is a person's unique preferred level of environmental stimulation (Berlyne, 1960). Humans engage in exploratory behavior in an attempt to adjust their perceived stimulation level to their desired optimal level (McReynolds, 1971). OSL is therefore a key construct underlying many types of exploratory behaviors (eg:information search, variety seeking, risk taking, etc,). The consensus finding is that persons with high OSLs participate in more exploratory behavior than persons with low OSLs.
Once OSL is achieved, sensory receptors open up and attention is focused (Howard and Sheth, 1969). From an advertiser's perspective, attention is potentially the most important index of consumer information search behavior (Goodwin, 1979). Research suggests that OSL may be a major determinant of consumer behavior with strong exploratory elements (Zuckerman, 1979).
Given the vastness of the information available on the World Wide Web, it is hypothesized that OSL with figure prominently in Web usage patterns.
This study will use innovative new methods to examine the Web user population. It is proposed that a Web site be constructed to survey Web users in their natural environment. This special sampling technique has been proven to be well suited to meet the new technology challenge by several pioneering researchers (Pitkow and Kehoe, 1995).
The proposed Web survey will function by residing on a dedicated server computer accessible to any Web user who becomes aware of its existence. Surveys residing on a Web-site feature short completion times and minimal respondent complexity, which yields increased response rates and decreased refusal rates (Pitkow etal., 1995) .
Web-based electronic survey research presents the classic threat to external validity : self-selection bias. A highly diversified publicity campaign is proposed to create awareness about the survey and reduce any systematic effects resulting from self-selection bias.
Reliable Web population demographic information is becoming available [Nielsen Media Research, (1995), O'Reilly & Associates, (1995)]. Therefore, respondent's demographic data can and will be compared to the known Web population to reveal self-selection and non-response problems.
Technological tools will be utilized to judge the uniqueness of responses. The content of multiple submissions from one positively identified computer can be compared for uniqueness, thereby providing some measure of control over individuals who deliberately induce self-selection bias.
Descriptive statistics and inferential statistics will be applied using the standard SPSS statistical package.
Chapter two provides an extensive review of of relevant literature. Chapter three details the researcher's proposed methodology. The proposed questionnaire is included in the appendix.
This chapter reviews the psychological literature pertinent to exploratory behavior theory and the underlying optimum stimulation level construct (OSL). Exploratory behavior and OSL will also be reviewed from the consumer information search behavior perspective. Literature relevant to the World Wide Web environment and it's users will also be reviewed. Finally, a rationale will be developed and hypotheses to be tested will be proposed.
Exploratory behavior has received considerable attention in the psychology literature. The concept of optimum stimulation level (OSL) was simultaneously introduced by Hebb(1955) and Leuba(1955) . The basic principle is that people prefer a certain level of environmental stimulation and that behavior will be motivated to attain a satisfactory level of stimulation.
Exploratory behavior is an attempt to regulate stimulation in order to achieve an individual's unique optimum stimulation level (Berlyne, 1960). Exploratory behavior is triggered by a change in a person's arousal state which is perceived to be less than optimal. Withdrawal from exploratory behavior results when the stimulation level greatly exceeds the preffered OSL. The character and complexity of stimuli are also important elements of the construct (Bettman, 1979).
Berlyne's Complexity Theory (1960) further posits that exploratory behavior can be classified as either specific or diversive behavior. The former refers to in-depth exploration of a single stimulus because of its arousal quality. The latter refers to non-directional behavior due to boredom.
Research confirms that individuals differences exist in the amount of stimulation perceived to be optimal (McReynolds, 1971). Furthermore, the consensus finding is that persons with high OSLs participate in more exploratory behavior than persons with low OSLs [(Zuckerman, (1979), Raju, (1981), Raju and Venkatesan (1980)].
Venkatesan (1973) and Raju (1977) conducted early reviews of exploratory behavior from the consumer behavior perspective. Sawyer (1977) and Faison (1977) linked advertising concepts to applications of exploratory consumer behavior. Raju (1980) investigated broad categories of exploratory behavior within the consumer context. Steenkamp and Baumgartner (1992) studied specific information search behaviors .The consensus finding is that OSL is positively correlated to many aspects of exploratory behavior within the consumer context.
In general, exploratory consumer behaviors can be categorized as either curiosity-motivated, variety seeking, or risk taking (Raju, 1980). Consistent with Berlyne (1960), consumer information search behavior is a form of curiousity motivated behavior which can be either goal-directed (ie:in purchase decision-making situations), or true exploratory where the acquisition of information is its own end (Steenkamp and Baumgartner, 1992).
Information search is an important component of consumer problem solving and decision making (Janis and Mann, 1977). Consumers also seek information to enhance decision quality or to reduce risk [Bettman, (1979), Pung and Staelin, (1993)].
The relationship between OSL and information search behavior has been confirmed by a number of researchers [Raju, (1980), Joachimsthaler and Lastovicka (1984)]. Price and Ridgway (1982) established the relationship of vicarious exploratory behavior and OSL. Exploratory behavior through shopping was also significantly correlated to OSL by Wahlers,etal. (1986).
The relationship between OSL, advertising stimuli, and the consumer consumption decision process is clear. "Once arousal is raised above the OSL, sensory receptors open and attention is thereby increased" (Howard and Sheth, 1969).
Advertising strategies are developed to grab and hold a consumer's attention in order to achieve a communication goal (eg:attitude change, attitude reinforcement, brand-switching, etc.). From an advertiser's perspective, attention is possibly the most important index of consumer information search behavior (Goodwin, 1979). Raju and Venkatesan (1980) have demonstrated the usefulness of including OSL in the study of consumer information search behavior.
A total of 37 million persons are estimated to have Internet access in the US and Canada. Approximately 18 million of these users have accessed the Web in the past three months. No reliable estimates are available for the global user population. The average Web user is affluent, white, male, and highly educated (Nieslen Interactive Services, 1996).
Other recent studies of the Web user population have similarly focused almost exclusively on discovering demographic and psychographic information about users. This information is useful when applying descriptive statistics, but provides an incomplete characterization of Web users.
Research suggests that OSL is a major determinant of consumer behavior with strong exploratory elements (Zuckerman, 1979). Given that the World Wide Web represent a vast and easily accessible wealth of information for a consumer to incorporate into their search behavior, it is hypothesized:
H1: As compared to individuals with lower OSLs, individuals with higher OSLs will report greater curiosity motivated information search behavior.
H2:As compared to individuals with higher OSLs, individuals with lower OSLs will report greater goal directed information search behavior.
Survey research design has traditionally offered a powerful, economical method for researchers interested in making inferences about the characteristics, behaviors, or attitudes of a population. It is proposed that a Web site be constructed to survey Web users in their natural environment.
The gross pool of respondents will be anyone who can access the World Wide Web which is estimated to be(insert stat here) worldwide. The researcher is primarily interested in Web users from the United States. These users have a well documented demographic profile which will be useful when conducting analytic procedures . Information about other users is sketchy at best. Therefore, the net pool of respondents is approximately 18 million people (Nielsen Media Research, 1995).
Respondents are limited to those people who are using the World Wide Web. This method is highly efficient at garnering respondents who are actual Web users. Other methods such as random-digit-dialing telephone sampling typically require the researcher to expend enormous effort to obtain a sample. For example, the CommerceNet/Nielsen Internet Demographics Survey placed 280,000 calls to obtain 4,000 completed surveys (Nielsen Media Research, 1995). In contrast, the GVU 4th World Wide Web User Survey netted responses from 23,000 Web users during the one-month period the survey resided on the Web (Pitkow and Kehoe, 1995).
A questionnaire has been constructed which contains three parts: questions about Web usage patterns, a standardized measure of OSL, and demographic questions. The questionnaire contains a total of twenty-three questions. Sixteen questions are related to demographics. Six questions are related to Web usage patterns.
The questionnaire also contains three additional questions which are not part of this proposed study. These questions have been placed in the questionnaire as a courtesy to other researchers who are also studying Web users.
Several measures of OSL have been widely utilized and tested by researchers of consumer behavior. The major instruments are the Arousal Seeking Tendency Scale (AST) of Mehrabian and Russell (1974) and later revised by Mehrabian (1978; AST-II), the Sensation Seeking Scale (SSS:the latest version is SSS-V) of Zuckerman, Kolin, Price, and Zoob (1964), the Change Seeker Index of Garlington and Shimota (1964), and the Novelty Experiencing Scale (NES) by Pearson (1970).
The AST instrument was selected for operationalizing OSL because it has been demonstrated to have a high reliability and validity. The original Kuder-Richardson reliabilty was found to be 0.87 and the test-retest reliability was found to be 0.88 (Mehrabian and Russell, 1974). The refined AST-II instrument features fewer questions and has demonstrated an even higher reliability of 0.93 (Mehrabian, 1978).
Constructs such as anxiety and extroversion have been correlated with the AST scale in order to establish construct validity (Ragu, 1980). Construct and convergent validity of AST-II was also been demonstrated when confirmatory factor analysis was applied to all four scales (Steenkamp and Baumgartner, 1992). This analysis confirmed the AST-II and CSI to the best operationalization of OSL.
The AST-II measure features thirty-two items scored on a nine-point Likert scale. The CSI is more time consuming to administer with almost three times as many items as the AST-II.
Berlyne's Complexity Theory (1960) posits that exploratory behavior can be classified as either specific or diversive behavior. The former refers to in-depth exploration of a single stimulus because of its arousal quality. The latter refers to non-directional behavior due to boredom.
In general, exploratory consumer behaviors can be categorized as either curiosity-motivated, variety seeking, or risk taking (Raju, 1980). Consistent with Berlyne, consumer information search behavior is a form of curiousity motivated behavior which can be either goal-directed (ie:in purchase decision-making situations), or true exploratory where the acquisition of information is its own end (Steenkamp and Baumgartner, 1992).
Raju's scale (1980) will be use to measure curiosity motivated exploratory information search on the Web. This measure has been significantly correlated to OSL by several researchers [Raju (1980), Joachimsthaler and Lastovicka (1984), Wahler etal. (1986)]. Raju also reported the scale to have a Spearman-Brown reliability coeficient of .761. The Raju scale features nine items scored on a 5-point Likert scale.
Two original items will be added to the Raju scale which will operationalize goal directed exploratory information search on the Web. Respondents will be asked to rate the following statements: "When I use the Web, I'm usually trying to find a specific site or piece of information", and "When I use the Web, I'm usually just surfing around to see what's out there".
The Web survey will function by residing on a dedicated server computer accessible to any Web user in the world who becomes aware of its existence. A respondent who chooses to participate will retrieve the survey from the server's database. Using a mouse, responses are easily completed and available for review before submission.
Because Hyper Text Transfer Protocol (HTTP) allows for the automatic logging of responses, respondents will be relieved of the burden of submitting the completed responses as the server will automatically record the data into its memory.
Surveys residing on a Web-site feature short completion times and minimal respondent complexity, which yields increased response rates and decreased refusal rates (Pitkow, Recker, & Gupta, 1994). A pilot study of the proposed study showed the questionnaire should take respondents about (insert stat here) minutes to complete. After completing the questionnaire respondents will be rewarded with a fractal display.
Electronic survey research poses unique methodological challenges for researchers attempting to characterize Web users. By design, surveys residing on a Web-site garner respondents only if they both become aware of the site and self-select themselves to participate. Thus, Web-based electronic survey research presents the classic threat to external validity : self-selection bias.
The sampling frame for a Web survey is all Web users. The inherent difficulty is getting a representative sample. Web surveys are accessible to any Web user anywhere. This loss of survey distribution control means Web users can self-select whether or not they choose to participate in a Web survey. Additionally, they can further self-select not to take the survey or quit before completing the survey.
Non-response bias can be severe if respondents don't know about the survey and therefore don't get an equal chance to participate. Promoting awareness of a Web survey is difficult because the Web lacks the capability to broadcast announcements to all users.
Web users also can skew data by taking a Web survey multiple times by faking different identifications each time. Additional bias can be introduced because the Web researcher is incapable of determining whether or not a respondent is truthful when completing responses.
The representativeness of a Web sample can be improved by getting as many unique visitors as possible to visit the Web survey site. This can be accomplished by diversifying the exposure of the site. Pitkow, et.al. (1995) argue a highly diversified publicity campaign should be designed to simulate a broadcast mechanism. For example, The 3rd GVU WWW User Survey ( Pitkow, et.al., 1995) was publicized via:
A publicity campaign similar in scope is proposed to reduce any systematic effects resulting from self-selection bias.
Other special techniques will allow for discovery of systematic sampling bias. The survey itself will contain questions regarding how the respondents found out about the survey. Respondents can then be stratified accordingly in order to discriminate systematic differences between these groups .
The problem of self-selection challenges the design of any survey. Anyone who decides whether or not to participate in a survey has self-selected themselves. Reliable Web user demographic information is beginning to become available [Find/SVP, (1995); O'Reilly & Associates (1995); Nielsen Media Research (1995); SRI International (1995)]. Therefore, the demographic characteristics of respondents can and will be compared to the known Web population characteristics to reveal self-selection problems.
Technological tools will be used to judge the uniqueness of electronic survey responses. Identification of the origin of a response can be established, recorded, and tracked because a respondent's hostname and IP address are passed into a Web server shell (Pitkow, Recker, & Gupta, 1994). This knowledge does not allow the researcher to precisely identify multiple submissions from the same respondent. However, the content of multiple submissions from one positively identified machine can be compared for uniqueness, thereby providing some measure of control over individuals who deliberately induce self-selection bias.
Descriptive statistics will be applied to the demographic information using the SPSS package. Inferential statistics will be applied to test hypotheses. Given the potentially large n-number involved, explained variance will be analyzed along with traditional significance tests.
A Growing Internet is Trying to Taker Care of Business. (1993, December
12). New York Times, p. F7.
Berlyne, D.E. (1960). Conflict, Arousal, and Curiosity. New York:McGraw
-Hill.
Berners-Lee, T.J. etal. (1992). The World Wide Web:The Information
Universe. Electronic Networking:Research, Applications, and Policy, Spring,
2(1), 52-8.
Bettman, J.R. (1979). An Information Processing Theory of Consumer
Choice. Reading, MA:Addison-Wesley Publishing Co.
Faison, E.W.J. (1977). The Neglected Variety Drive:A Useful Concept
for Consumer Behavior. The Journal of Consumer Research, 4, 172-5.
Find/SVP (1995). The American Internet User Survey (URL=http://www.findsvp.com).
Garlington, W. K., and Shimota, H. E. (1964). The Change Seeker
Index:A Measure of the Need for Variable Stimulus Input. Psychological
Reports, 14, 919-924.
Godwin, S.A. (1979). Impact of Stimulus Variables on Exploratory
Behavior. Advances in Consumer Research, ed. Jerry C. Olson, Vol.7, 264-69
(Ann Arbor:Association for Consumer Research.
Hebb, D.O. (1955). Drives and the C.N.S.(Central Nervous System).
Psychological Review, 62-243-54.
Hoffman, D.L. and Novak, T.P. (1995). Marketing in Hypermedia Computer-Mediated
Environments:Conceptual Foundations (URL=http://www2000.ogsm.vanderbilt.edu).
Howard, J.A., and Sheth, J.N. (1969). The Theory of Buyer Behavior.
New York:Wiley.
Jannis, I.L., and Mann, L. (1977). Decision Making. New York:The
Free Press.
Joachimsthaler, E.A., and Lastovicka, J.L. (1984). Optimum Stimulation
Level-Exploratory Behavior Models. Journal of Consumer Research, 11, 830-35.
Leuba, C. (1955). Toward Some Integration of Learning Theories:The
Concept of Optimal Stimulation. Psychological Reports, 1, 27-33.
McReynolds, P. (1971). The Nature and Assessment of Intrinsic Motivation,
Advances in Psychological Assessment. Vol.2 ed. Paul McReynolds, Palo Alto,
CA:Science and Behavior Books, 155-177.
Mehrabian, A., and Russell, J. A. (1974). An Approach to Environmental
Psychology, Cambridge, MA:MIT Press.
Mehrabian, Albert (1978). Characteristic Individual Reactions to
Prefered and Unpreferred Environments. Journal of Personality, 46 (December),
717-731.
Nielsen Media Research (1995). The CommerceNet/Nielsen Internet
Demographics Survey (URL=http://www.nielsenmedia.com/demo.htm). survey-paper.html).
O'Reilly & Associates (1995). Defining the Internet Opportunity.
(URL=http:www.ora.com/survey).
Pearson, P. H. (1970). Relationships between Global and Specified
Measures of Novelty Seeking. Journal of Consulting and Clinical Psychology,
43 (2), 199-204.
Pitkow, J.E. and Recker, M.M. (1994). Results from the First World
Wide Web User Survey (URL=http://www.cc.gatech.edu/pitkow/survey-1-1994/
survey-paper.html).
Pitkow, J.E., Recker, M.M., and Gupta, S. (1994). The GVU Center's
2nd WWW User Survey (URL=http://www.cc.gatech.edu/gvu/user_surveys/ survey-09-1994/html-paper/survey_2_paper.html).
Pitkow, J.E., et. al. (1995). The GVU Center's 3rd WWW User Survey
(URL=http://www.cc.gatech.edu/gvu/user_surveys/ User_Survey_Home.html).
Pitkow, J.E. and Kehoe, C. (1995). The GVU Center's 4th WWW User
Survey (URL=http://www.cc.gatech.edu/gvu/user_surveys/survey-10- 1995/).
Price, L.L., and Ridgway, N.M. (1982). Use Innovativeness, Vicarious
Exploration and Purchase Exploration:Three Facets of Consumer Varied Behavior.
1982 Educator's Conference Proceedings, ed. Bruce J. Walker etal., Chicago:American
Marketing Association, 56-60.
Pung, G.N., and Staelin, R. (1983). A Model of Consumer Information
Search for New Automobiles. Journal of Consumer Research, 9, 366-380.
Raju, P.S. (1977). Theoretical Perspectives on Exploratory Behavior:A
Review and Examination of Their Relevance for Consumer Research. paper
#67, Working Series in Marketing Research, The Pennsylvania State University,
PA.
Raju, P.S. (1980). Optimum Stimulation Level:Its Relationship to
Personality, Demographics, and Exploratory Behavior. Journal of Consumer
Research, 7, 272-82.
Raju, P.S. (1981). Theories of Exploratory Behavior:Review and Consumer
Research Implications. Research in Marketing, Vol. 4, ed. Jagdish N. Sheth,
Greenwich, CT:JAI, 223-249.
Raju, P.S. and Venkatesan, M. (1980). Exploratory Behavior in the
Consumer Context:A State of the Art Review. Advances in Consumer Research,
Vol. 7, ed. Jerry C. Olson, (Ann Arbor, MI:Association for Consumer Research),
258-263.
Sawyer, A.G. (1977). Repetitiion and Affect:REcent Empirical and
Theoretical Developments. The Journal of Consumer Research, 3, 95-100.
SRI International (1995). Exploring the World Wide Web Population's
Other Half (URL=http://future.sri.com/vals/vals- survey.results.html).
Steenkamp,J-B.E.M. and Baumgartner, H. (1992). The Role of Optimum
Stimulation Level in Exploratory Consumer Behavior. Journal of Consumer
Research, 19, 434-448.
Venkatesan, M. (1973). Cognitive Consistency and Novelty Seeking.
Consumer Behavior:Behavioral Sources, eds., S. Ward and T.S. Robertson,
Englewood Cliffs, NJ:Prentice-Hall, 354-84.
Whalers, R.G., etal. (1986). The Congruence of Alternative OSL Measures
with Consumer Exploratory Behavior Tendencies. Advances in Consumer Research,
Vol. 13, ed.Richard J. Lutz, Provo, UT:Association for Consumer Research,
398-402.
Zuckerman, M., etal. (1964). Development of a Sensation Seeking
Scale. Journal of Consulting Psychology, 28, 477-82.
Zuckerman, M. (1979). Sensation Seeking:Beyond the Optimal level
of Arousal. Hillsdale,NJ;Erlbaum.
Return to Proposals Return to Course Notes Return to Syllabus