Student Internet Purchases - Wiley Online Library

11 downloads 1716 Views 86KB Size Report
online purchasing behavior of this consumer market could yield insight relevant to marketers ... analyze online buying behavior among college students.
Student Internet Purchases Pamela S. Norum University of Missouri–Columbia

Since the onset of the Internet shopping format, e-commerce sales have soared, and Generation Y consumers, the first generation to grow up with the Internet, have turned into young adults. As part of Generation Y, college students represent a lucrative market for businesses selling a wide array of goods and services, and they are extremely computer savvy. The purpose of this study was to empirically analyze online buying behavior among college students. Based on an economic framework, factors affecting the online purchase of nine different merchandise categories were examined. Data were analyzed from 4,688 students enrolled in a major Midwestern university. Logistic analysis revealed that variations in the effect of age, gender, income, car ownership, ability to identify a secure Internet site, and compulsive buying behavior existed between the merchandise categories. This study has implications for educators, marketers, and policy makers. Keywords: Internet buying; Generation Y; purchasing patterns

E-commerce sales more than tripled between 2000 and 2006, reaching a record high of more than 114 billion dollars in 2006 (U.S. Department of Commerce, 2007). At its onset, many questions arose about the effect that the Internet would have on consumer buying behavior. Who would shop on the Internet? Would the Internet lead to the demise of brick-and-mortar stores? Although the Internet appears to have its place alongside other retail formats, there is still much to know about online purchasing behavior. In fact, with changes in consumer attitudes toward technology, as well as demographic shifts, characteristics of online purchasers today are different from just a few years ago. Although many members of Generation Y would have been too young just 10 years ago to make significant purchasing decisions, this generation is now having an effect on American industries ranging from automobiles and technology to fashion and finance (Weiss, 2003). This generation, also known as the Millenial generation, is distinctive for several reasons, including its size, financial means, and technological savvy. Seventy-two million Americans, born between 1977 and 1994 (Wellner, 2000), make up the Millennial generation, making it almost as large as the baby boom generation (Weiss, 2003; Wellner, 2000). This market spends $187 billion annually (Weiss, 2003), and the spending power of this generation will continue to increase as they move through the different stages of adulthood. In addition to its sheer size, the Millenial generation has access to more credit than any generation before it (Weiss, 2003). It is the first “high-tech” generation (Mitchell, 1998), having been brought up with computers as an everyday part of their lives. Many individuals within this generation are technological innovators. In general, technological innovators have been found to be younger and more educated than the general population, and the opinions of these early technology adopters Author’s Note: Pamela S. Norum, PhD, is an associate professor in the Department of Textile and Apparel Management. Family and Consumer Sciences Research Journal, Vol. 36, No. 4, June 2008 373-388 DOI: 10.1177/1077727X08318705 © 2008 American Association of Family and Consumer Sciences

373

374

FAMILY AND CONSUMER SCIENCES RESEARCH JOURNAL

eventually filter down to other consumer segments (Watchravesringkan & Shim, 2003). This pattern suggests that insight into changes in technological trends might be gained by analyzing the behavior of younger consumers who are focused on attaining higher levels of education. In 2004, individuals in Generation Y were 10- to 27 years old and, thus, encompassed students of traditional college age. College students represent a lucrative market for businesses selling a wide array of goods and services. “College students as a whole represent a multibillion dollar market: In 2002, 5.9 million full-time students in four-year institutions spent some $9.2 billion during the school year on discretionary items, up 27 percent from 1997” (Yin, 2003, p. 20). Members of this generation who are attending college have daily exposure and access to computers, as well as the Internet. Paralleling the birth of the Millenial generation was the development of the Internet. Although the Internet has a multitude of uses, it does provide a nonstore format for shopping. Given the technological sophistication and purchasing power of college students, examination of online purchasing behavior of this consumer market could yield insight relevant to marketers, educators, and policy makers alike. Answers to the following questions could be insightful: Is online buying pervasive among college students? In what ways can the purchasers be distinguished from the nonpurchasers? Among the Internet purchasers, are there differences in specific goods or services that they buy? Does credit play a role in purchasing? What are other factors that influence online buying behavior? With these considerations in mind, the purpose of this study was to empirically analyze online buying behavior among college students. This research will go beyond previous studies that looked at only one product category, such as apparel, or that looked at Internet shopping in general (and not by specific product categories). This study will also focus on Generation Y consumers, a specific market segment that is more technologically oriented than prior generations. Because education tends to be positively related to Internet purchasing behavior, this study will further focus on Millenials enrolled in college. In addition, the role of gender will be explored. One way to identify potential differences in online shopping between men and women is to include a variety of product categories in the analysis. In addition, other factors that may influence Internet buying will also be examined. These include Internet security, compulsive buying behavior, credit usage, and retail accessibility. The results of the study should be of interest to retailers interested in both buying behavior and security concerns; family and consumer sciences professionals who provide education or counseling with regard to consumer purchase decisions and buying behavior; educators interested in consumer behavior, including issues related to Internet purchasing, compulsive buying behavior, security concerns, and credit card usage among college students; and policy makers interested in issues related to Internet security and credit card usage.

REVIEW OF LITERATURE

Previous researchers have examined consumer demand or purchasing behavior for a wide range of goods (Abdel-Ghany & Silver, 1998; Dyer, Burnsed, & Dyer, 2006; Norum, 2003; Paulin, 2001). A common thread among all of these studies was the use of an economic framework for analyzing consumer demand. From

Norum / INTERNET BUYING

375

that perspective, a key variable included in each analysis was a measure of income. Consumers’ ability to buy a particular good, the amount they consume, and/or the amount they spend is directly related to their income. Each of these studies has found a positive relationship between income and consumer demand. An economic framework will be used for this study as well. The relationship between online shopping behavior and a variety of demographic variables has been examined in previous studies. Although differences in age, income, and occupation were not found between three categories of shoppers (Internet only, mall and Internet, and mall only), significant differences were found for education and gender. Internet shoppers tended to be more highly educated (Shim, Eastlick, & Lotz, 2000). This coincides with previous studies that found higher education to be positively related to online shopping (Shim & Drake, 1990; Shim et al., 2000). With respect to gender, men were more likely to be Internet only, or Internet and mall shoppers, than women. This was true for both types of product categories (cognitive and sensory) examined in the analysis (Shim et al., 2000). Goldsmith and Goldsmith (2002), using a college student sample, found no statistically significant relationship between age and gender, and online purchasers and nonpurchasers of apparel. In another study, female respondents without a college degree had a more positive attitude about online shopping compared with women with a 4-year degree (Watchravesringkan & Shim, 2003). They also had a more positive attitude than men with or without a college degree. Women were also more likely than men to use the Internet to search for apparelrelated information (Watchravesringkan & Shim, 2003). Early research on Internet usage indicated that men were heavier users than women. However, it is possible that, over time, shifts have occurred between men and women and their use of the Internet. It is estimated that women will account for 52% of total e-commerce sales in the United States in 2007 (Internet News, 2003). It is also possible that differences may also vary depending on the product category being purchased. However, this is an area that needs further investigation. Using an online survey, Taylor and Cosenza (2000) compared the behavior of women who purchased women’s clothing both online and in malls with women who did not buy online. It was found that Internet experience influences purchase behavior. That is, prior online purchase of products, in general, increased the likelihood of buying women’s apparel online. Factors related to the perceived risk of shopping for women’s clothing on the Internet were also found to have a significant effect on online purchasing behavior. In this study, the risks were related to product evaluation rather than security of the Internet itself. However, this latter issue is one that has been identified as a major reason that consumers refrain from buying on the Internet (Kwon & Lee, 2003; Watchravesringkan & Shim, 2003). Because concerns with regard to payment security on the Internet are a deterrent to online purchases, some researchers have specifically addressed this issue. Analysis of computer users in 15 metropolitan cities found that consumers’ attitudes about secure transactions were related to Internet shopping intention (Watchravesringkan & Shim, 2003). Kwon and Lee (2003) found that payment security concerns directly influence online purchases. The greater the concern, the less likely a consumer was to purchase on the Internet. Offline payment options, and dispelling consumers’ fears concerning the use of their credit card online, were two suggestions made by the authors. The latter recommendation was suggested in light of the fact that payment security does not appear to be any greater

376

FAMILY AND CONSUMER SCIENCES RESEARCH JOURNAL

risk online than in a brick-and-mortar store (Rowley & Okelberry, 2000). A related area, but one that has not been addressed in previous research, is whether or not a consumer knows how to recognize a secure site. Having this knowledge would, presumably, reduce consumer concerns related to online purchases. Phau and Lo (2004) investigated differences between fashion innovators and noninnovators with respect to impulse consumption and Internet purchases for fashion items. An underlying idea was that the Internet provides consumers with a channel for immediate impulse buying. The results indicated that fashion innovators show more impulsive buying behavior than noninnovators, but no differences were found with respect to Internet purchases. Although Phau and Lo (2004) focused on impulsive buying, Lee, Lennon, and Rudd (2000) examined the relationship between compulsive consumption and one type of electronic shopping, specifically television shopping. Although impulse buying and compulsive buying are both types of unplanned purchases, compulsive buying tends to have more negative consequences (O’Guinn & Faber, 1989). Using a national sample of television shoppers, Lee et al. (2000) estimated that 10% of the sample (n = 334) had compulsive buying tendencies, as measured by the Compulsive Buying Scale (CBS) developed by Faber and O’Guinn (1992). Compulsive buying was positively related to exposure to television shopping channels. Because the Internet provides another type of electronic shopping format, the researchers suggested studying compulsive buying within the context of Internet shopping. Both formats are electronic and private and encourage the use of credit cards. In light of this latter characteristic, it should be noted that attitudes toward credit card use have been linked to compulsive buying (Roberts, 1998). Roberts found compulsive buying to be related to irrational credit card use, gender, self-esteem, shopping frequency, perceived social status associated with buying, and television viewing among a sample of college students. Roberts and Jones (2001) found irrational credit card usage to play a moderating role in compulsive buying behavior in a sample of college students. There could be some value in looking at both compulsive buying and irrational credit card usage within the context of Internet shopping. Shoham and Brencic (2003) investigated the relationship between gender, a consumer’s tendency to make unplanned purchases, and a consumer’s tendency to buy products not on shopping lists. Based on a sample of Israeli consumers, women were more likely to display tendencies toward compulsivity. Unplanned purchases and buying items not on a shopping list were both positively, and significantly, related to compulsive buying. It would seem that the Internet would provide plenty of opportunities for making unplanned purchases, whether an individual is engaged in recreational activities like surfing the Web, or visiting retail sites to gather information. This further supports investigating the relationship between compulsive buying and Internet purchases. Using a convenience sample of 305 undergraduate students, Yurchisin and Johnson (2004) examined the influence of apparel-product involvement, and several other factors, on compulsive buying. Fifteen percent of the sample were classified as compulsive buyers based on the scale developed by Edwards (1993). Perceived social status and materialism had positive and significant effects on compulsive buying, whereas self-esteem had a negative and significant effect on compulsive buying. The result of greatest interest was that apparel-product involvement was positively and significantly related to compulsive buying. That

Norum / INTERNET BUYING

377

is, the more importance people placed on apparel, and the greater their interest in apparel, the more likely they were to be compulsive buyers. This suggests that compulsive buying can be expected to have a significant influence on online apparel purchases, in particular. Previous research has found that accessibility to retail venues influences shopping behavior (Spiggle & Sewall, 1987). Various factors are related to accessibility including distance, types of stores, and number of stores (Spiggle & Sewall, 1987; Taylor & Cosenza, 2000). Generally, it has been assumed that consumers have access (through transportation) to traditional retail formats (e.g., stores at malls). With the development of the Internet and online shopping, accessibility was viewed from a different perspective. Instead of being concerned about whether a consumer had transportation available to access retailers, access to a computer became a consideration (Shim et al., 2000; Taylor & Cosenza, 2000). In fact, some studies focused on Internet shopping have selected their sample based on computer ownership and Internet access (Kim, Kim, & Kumar, 2003). For the college student sample, however, computer access is not a major obstacle. Students do have 24-hour access to public computers, even if they do not own their own personal computer. Computers are widely available on campus in computing labs and dormitories. However, the assumption that college students have transportation readily available so they can access traditional retail formats may be misguided. Not all students have easy access to malls or other retail outlets because they do not have their own transportation. Thus, having a car with them at college, as a measure of retail accessibility, will be included in this study.

THEORETICAL MODEL Consumer Demand Theory

Consumer demand theory provided the theoretical framework for this study. From this perspective, consumers maximize their utility subject to their budget constraint from which demand functions for various goods and services were derived (Varian, 1999). A consumer’s utility function was defined as U = u (X1, XAOG)

where X1 = good 1 and XAOG = all other goods. A consumer maximizes utility subject to its budget constraint: I = P1 X1 + PAOG XAOG

where I = total income, P1 = price of clothing, and PAOG = price of all other goods. Maximization of the utility function subject to the income constraint yielded the demand function for good 1: Q1 = f (P1, PAOG, I; T)

where Q1 = quantity demanded of good 1 and T = tastes and preferences. The quantity demanded of a good is a function of income, prices, and tastes and preferences (Varian, 1999). When using cross-sectional data, as in this study, prices were assumed to be constant across consumers, over the time period of the

378

FAMILY AND CONSUMER SCIENCES RESEARCH JOURNAL

analysis. Therefore, prices were suppressed in the empirical equation. The empirical equation included measures for income and factors to control for tastes and preferences. Selected variables found to influence Internet buying in previous studies were included to measure tastes and preferences.

METHOD Empirical Model and Analysis

For this study, information is available on whether or not a consumer is an online purchaser of at least one category of merchandise, but neither the exact quantity nor expenditure amount is known. It was known, however, that the consumer did purchase some nonzero quantity. As a consequence, the dependent variable was a dichotomous variable, and logistic regression analysis was an appropriate statistical technique (Tacq, 1997). The general form of the equation estimated in this study is IP = a + b1I + b2A + b3G + b4SS + b5CB + b6IC + b7CO + e

where IP = Internet purchase; I = income; A = age; G = gender; SS = secure site; CB = compulsive buying behavior; IC = irrational credit card usage; CO = car ownership; a = the intercept; bi = regression coefficient; and e = error term. The coefficients produced by the logistic procedure cannot be interpreted in the same way as the regression coefficients from ordinary least squares regression. However, the odds ratio provided coefficients that represent the effect of changes in the independent variables on the dependent variable. In this context, the coefficient on a dummy variable can be interpreted as a percentage difference relative to the comparison variable. For example, an odds ratio of 1.98 on gender in the clothes purchasing equation indicates that women are 98% more likely to buy apparel online than men (which, as the comparison category, would have a value of 100). On the other hand, a coefficient of .42 on gender in the video purchasing equation indicates that women are 58% less likely to buy videos online than men. Sample and Data Collection

To collect data for this study, students enrolled in a major university located in a medium-sized Midwestern town were asked to complete a Web-based questionnaire. To distribute the questionnaire, both graduate and undergraduate students (N = 27,003) were contacted by e-mail through the university system. The e-mail asked students to participate in the study. Interested students clicked on a Web address to be able to complete the survey. Two additional follow up e-mails were sent as reminders 10 and 20 days later. Students were offered the opportunity to enter their name in a drawing for one of three $150 gift certificates, as an incentive to participate in the study. The response rate was 27.19%, with a total of 7,342 questionnaires returned. However, this study specifically focused on the 4,688 undergraduate students aged 18 to 27 (these ages correspond to the ages of the Millenial generation in 2004, the year the data were collected). Graduate students were not included in this study because they tend to be financially independent from their parents and have their own families, which makes them distinctly different from undergraduates. To be included in the analyses, the students had to provide complete information on the variables of interest.

Norum / INTERNET BUYING

379

Dependent Variable

Internet buying. The respondents were asked, “Do you make purchases on the Internet?” “Yes” and “no” were the response options provided. This variable was coded 1 if they did currently make purchases on the Internet, and 0 otherwise. This was the dependent variable in the empirical equation that was estimated for the whole sample. Those respondents who made purchases on the Internet were also asked to indicate the types of goods or services they buy online. This was done by asking the respondents to check all items that they buy on the Internet, from a list that was provided to them. Thus, although we do not know the exact quantity purchased, we do know that some nonzero amount was purchased. Therefore, dummy variables were created for the following purchase categories: clothing and accessories, entertainment (concert tickets, movies, dance clubs, etc.), travel (airfare, hotel, rental car), computer and related supplies, general books (other than textbooks), textbooks/school supplies, electronic equipment, and cosmetics and other personal items. The variable was coded 1 if a purchase was made in that category, and 0 otherwise. Independent Variables

Parents’ income. Economic factors, and income in particular, greatly influence purchasing behavior. To capture this effect, parental income was included in the empirical equation because students are frequently dependent on their parents for support. Dummy variables were created with the following categories: $25,00049,999; $50,000-74,999; $75,000-99,999; and $100,000 or more. The omitted category was $25,000 or less. Students who said they “didn’t know” their parents’ income were not included in the analysis. Based on economic theory, income was expected to be positively related to Internet purchases. As income increases, the demand for normal goods increases. All goods included in the analysis were expected to be normal goods. Age. Age has been found to influence consumer purchases. For this study, it was hypothesized that the younger Millenials will be more likely to buy online because younger people are expected to be more technologically oriented and tend to be technological innovators (Watchravesringkan & Shim, 2003). Gender. A dummy variable for gender was created with women equal to 1, and men equal to 0. Women tend to shop more than men, in general, and recent studies indicate that women seem to have a greater propensity toward Internet buying than men (Internet News, 2003; Watchravesringkan & Shim, 2003). Thus, it was hypothesized that women are more likely to buy on the Internet than men. However, when analyzing specific categories, variations are expected to exist. Within specific categories, women are expected to be more likely to purchase clothing (Yurchisin & Johnson, 2004) and makeup, which mirrors shopping patterns in nonelectronic formats. For the other categories, a specific direction was not hypothesized. Secure site. Concerns with regard to security negatively affect a consumer’s intention to purchase on the Internet (Kwon & Lee, 2003; Watchravesringkan & Shim, 2003). Thus, it was expected that a consumer who is able to identify a secure site

380

FAMILY AND CONSUMER SCIENCES RESEARCH JOURNAL

would be more willing to buy online compared with a consumer who cannot identify a secure site. This was a self-reported measure where respondents were asked, “Do you know how to recognize a secure site?” This variable was coded 1 if the respondent said he or she knew how to identify a secure site, and 0 otherwise. Compulsive buying. Kwak, Zinkhan, and Crask (2003) have suggested that there could be possible addictive behaviors, such as compulsive buying, associated with Internet shopping. Also, the Internet provides access to shopping 24/7, providing the compulsive buyer with more opportunities to buy. Phau and Lo (2004) pointed out that “since consumers practice impulse shopping in brick and mortar stores, they will also be likely to consume impulsively through the Internet” (p. 404). The same logic can be applied to compulsive buying. Compulsive buying was expected to be positively associated with Internet purchases. The Diagnostic Screener for Compulsive Buying (DSCB; Faber & O’Guinn, 1992) has been used by a number of previous researchers (Kwak et al., 2003; Lee et al., 2000; Roberts, 1998; Roberts & Jones, 2001) for both college-age and adult samples and was used in this study. In this study, for ease of interpretation, the scores were multiplied by –1 so that a higher score indicated a greater propensity toward compulsive buying. Faber and O’Guinn (1992) provided evidence for the reliability of their scale (α = .95) as well as support for the validity of their scale. Roberts and Jones reported a Cronbach’s alpha equal to .78 for the DSCB. Alpha for the DSCB was .78 for this study. It was hypothesized that the higher the score on the DSCB, the more likely a person would be to make purchases on the Internet. Irrational credit card usage. Attitudes toward credit cards may have a direct effect on Internet purchases and, therefore, were included in this study. Roberts and Jones (2001) developed a measure of irrational credit card usage based on 12 different items. For example, respondents indicated the extent to which they felt they had too many credit cards, had their credit cards at their maximum limit, and worried about how they would pay off their credit cards. The items were measured on a 5-point Likert-type scale. An overall score on credit card usage was obtained by summing the responses to the individual items, except where reverse coded. The total score, in this study, was reversed so that a higher score indicates more irrational credit card usage. Roberts and Jones reported an alpha equal to .81, indicating good internal consistency. The scale used in this study had a coefficient of α = .80. Because the opportunity to shop using a credit card is available 24 hours a day on the Internet, it is possible that students with more irrational credit card behaviors will be more likely to buy on the Internet compared with students with more rational credit card behaviors. Thus, it was hypothesized that irrational credit usage will be positively associated with Internet buying. Car ownership. Accessibility to retail channels influences shopping behavior (Shim et al., 2000; Spiggle & Sewall, 1987; Taylor & Cosenza, 2000). Although in some studies, this referred to computer access to the Internet, it can also refer to access to transportation that provides access to physical shopping outlets. A dummy variable for whether a student had a car at school (1) or not (0) was included in the analysis. It was hypothesized that if a student does not have access to a car, he or she will be more likely to purchase goods on the Internet. This relationship may not hold true for services, such as travel.

Norum / INTERNET BUYING TABLE 1:

381

Logistic Regression, Internet Purchasing, Dependent (n = 4,599)

Variable Intercept $25,000-49,999 $50,000-74,999 $75,000-99,000 $100,000 or more Age Gender Recognize secure site Compulsive buying Irrational credit card use Car available

Parameter Estimate

Standard Error

Odds Ratio Estimate

–4.67*** 0.11 0.18 0.18 0.50*** 0.26*** –0.04 1.09*** 0.12*** –0.01** 0.32***

0.52 0.15 0.15 0.15 0.15 0.02 0.09 0.08 0.03 0.006 0.10

1.12 1.20 1.20 1.65 1.3 0.96 2.97 1.12 0.99 1.37

NOTE: X 2 = 477.87.*** *p < .10. **p < .05. ***p < .01.

DESCRIPTIVE STATISTICS

The average age of the college student sample was 20.69 years. More than two thirds (68.41%) of the sample were female (compared with the 52.4% of the actual student body at the time of data collection). Both male and female students were given an equal opportunity to participate in the study. It is possible that the incentive for participating (a drawing for one of three $150 gift certificates to a local mall) had a greater effect on the women than men. Fifty percent of the sample came from households in which the parental income was $75,000 or more, and 69.94% of the students had at least one credit card. Seventy-eight percent of the sample indicated that they currently bought on the Internet, whereas slightly fewer (70.94%) said they knew how to recognize a secure site. Based on the cutoff point for the DCBS, suggested by Faber and O’Guinn (1992), 7.76% could be classified as compulsive buyers. The irrational credit card usage score ranged from 12 to 55, with a mean equal to 24.78. Eighty-five percent of the sample had a car with them at school.

RESULTS

For the full sample (n = 4,688) of both undergraduate purchasers and nonpurchasers, a logistic regression equation was run in which the dependent variable was whether or not an individual currently made purchases on the Internet. Sixteen students did not provide a response for this variable and could not be included in the analysis. The significant variables in the equation were age, parental income of more than $100,000, availability of a car, credit card use, ability to recognize a secure site, and compulsive buying (see Table 1). There was a positive relationship between age and online buying. Because younger consumers are expected to be more technologically savvy, they were expected to be more likely to buy on the Internet. However, this hypothesis was not supported. Older students were 30% more likely to make a purchase online than younger students.

382

FAMILY AND CONSUMER SCIENCES RESEARCH JOURNAL

Income was hypothesized to be positively related to Internet purchases. Students from households with an income of $100,000 or greater were more likely to buy on the Internet than students from households with an income of less than $25,000. Car ownership was positively related to Internet buying. It was hypothesized that car ownership and the purchase of goods (versus services) would be negatively related, but that was not verified in this analysis. Irrational credit card use was negatively associated with Internet purchases. The odds ratio of .99 indicates that students with higher irrational credit use attitudes were 1% less likely to buy on the Internet. This result, although statistically significant, does not seem very substantial. The ability to recognize a secure site and compulsive buying were positively related to Internet purchasing, as hypothesized. To further analyze the online purchases of Millenials, specific merchandise categories were analyzed (see Table 2). These equations were estimated only for those students who indicated that they had made an online purchase. The total number of students who reported making one or more purchases on the Internet was equal to 3,645 (78.00%). However, not all of these purchasers made a purchase in every category. Thus, the sample size used for each estimated equation varied by product category. The results indicated tremendous variation across the merchandise categories with respect to each of the product categories. It is interesting that gender is the one variable that was significant in all of the equations whereas irrational credit card use was rarely significant. Parental income also showed mixed results by product category. Respondents with parents in the higher income brackets were more likely to buy clothing, entertainment, travel, and computers than respondents whose parents’ income was less than $25,000. Respondents from the higher income families were less likely to purchase textbooks and electronics online. It appears that the items being bought by respondents from the higher income families are more luxury goods that their parents may be willing and able to pay for. Age was negatively related to online purchases of clothing and entertainment, whereas it was positively related to online purchases of travel, computers, general books, and textbooks. (Younger students were more likely to buy clothes and entertainment and less likely to buy travel, computers, general books, and textbooks, relative to older students.) Thus, age does affect Internet buying, even among consumers within the same generational cohort, but the effect varies by product category. Prior research on the baby boomer generation indicated that age differences existed within that generation for expenditures on clothing (Norum, 2003). The differences varied by clothing category. For example, older boomers spent more on apparel for women and infants, whereas younger boomers spent more money on apparel for boys and girls. Gender demonstrated even wider variation among merchandise categories than did age. Women were more likely to buy clothing and accessories, travel, general books, textbooks, and makeup. Men were more likely to buy entertainment, videos/DVDs, computers, and electronics. In general, these results coincide with gender differences found in traditional shopping venues. The ability to recognize a secure Web site was positively related to all six merchandise categories in which this variable was significant. These categories included entertainment, videos/DVDs, computers, general books, textbooks, and electronics.

383

–1.33** 0.18 0.26* 0.29** 0.55*** –0.09*** 0.68*** 0.06 0.20*** –0.00 –0.02

6.69*** 0.27*** 0.61*** 0.21 0.29** 0.30** 0.63*** –0.09 –0.001 0.08 0.04

Intercept $25,000-49,999 $50,000-74,999 75,000-99,999 > $100,000 Age Gender Secure site Compulsive buying Irrational credit card use Own a car

Intercept Age Gender $25,000-49,999 $50,000-74,999 $75,000-99,999 > $100,000 Own a car Credit card use Secure site Compulsive buying

Clothing (n = 1,856)

X2 = 339.29*** Textbooks (n = 1,100)

1.31 1.84 1.24 1.34 1.35 1.88 0.92 1.00 1.08 1.04

1.20 1.30 1.34 1.73 0.92 1.98 1.06 1.22 1.00 0.88

Odds Ratio Estimate

X2 = 254.65*** Travel (n = 1,564)

Parameter Estimate

1.77*** 0.08*** –1.39*** 0.09 –0.21 –0.12 –0.26* –0.29** –0.005 0.64*** 0.006

0.50 0.33** 0.45*** 0.50*** 0.59*** –0.06*** –0.40*** 0.27*** 0.12*** 0.002 0.21**

X2 = 447.30*** Electronics (n = 823)

X2 = 98.21*** Computer (n = 1,048) 1.09 0.25 0.91 0.81 0.88 0.78 0.75 1.00 1.89 1.00

1.38 1.56 1.64 1.80 0.95 0.67 1.30 1.12 1.002 1.24

Odds Ratio Estimate

Entertainment (n = 1,634)

Parameter Estimate

Logistic Regression, Internet Purchases by Merchandise Category, Dependent

Variable

TABLE 2:

–4.71*** 0.18*** 0.19** 0.15 –0.24 –0.23 –0.11 –0.42*** –0.005 0.53*** –0.02

0.34 0.10 0.17 0.23 0.21 –0.032 –0.86*** 0.52*** 0.14*** –0.004 –0.19*

Videos (n = 1,448)

(continued)

X2 = 180.49*** Makeup (n = 504)

1.20 1.21 1.16 0.79 0.80 0.90 0.66 1.00 1.70 0.98

1.10 1.19 1.26 1.24 0.98 0.42 1.68 1.15 1.00 0.83

Odds Ratio Estimate

X2 = 205.41*** General Books (n = 1,073)

Parameter Estimate

384

*p < .10. **p < .05. ***p < .01.

–5.28*** 0.24*** 0.26*** –0.22 –0.53*** –0.57*** –0.76*** –0.31*** –0.01* 0.22*** –0.06*

Intercept Age Gender $25,000-49,999 $50,000-74,999 $75,000-99,999 > $100,000 Own a car Credit card use Secure site Compulsive buying 1.27 1.30 0.80 0.59 0.56 0.47 0.73 0.99 1.24 0.94

Odds Ratio Estimate

X2 = 286.84***

Parameter Estimate

(continued)

Variable

TABLE 2:

–0.62 0.02 –1.97*** –0.43** –0.28 –0.31* –0.30* –0.27** 0.007 0.60*** 0.06* 1.02 0.14 0.65 0.76 0.73 0.74 0.77 1.007 1.83 1.07

Odds Ratio Estimate

X2 = 609.95***

Parameter Estimate

–2.42*** 0.01 1.32*** –0.29 –0.07 –0.28 0.15 –0.42*** –0.003 0.08 0.23***

1.01 3.76 0.75 0.94 0.75 1.12 0.66 1.00 1.09 1.25

Odds Ratio Estimate

X2 = 219.20***

Parameter Estimate

Norum / INTERNET BUYING

385

Compulsive buying was also positively related to all merchandise categories in which it was significant, except for one. The higher the respondents’ compulsive buying score, the more likely they were to buy clothes, entertainment, videos/DVDs, electronics, and makeup. Respondents with higher compulsive buying scores were less likely to buy textbooks online, the one category that represents the greatest necessity among the categories analyzed. Irrational credit card use was significant only in the equation for textbooks and not in any of the other purchase categories. The more irrational the respondents’ credit card use, the less likely they were to buy textbooks online. Or conversely, the more rational the students’ credit card use, the more likely they were to buy textbooks online. It is possible that these latter students are more savvy consumers and realize that purchasing textbooks online can save money. Most notable about this variable is the lack of significance in all other equations. The result for the general online purchasing equation indicated that car ownership was positively related to the likelihood of making a purchase on the Internet. A closer look at the specific product categories reveals that this same relationship held true for entertainment, but not all other product categories. Respondents who had a car at school were significantly less likely to make online purchases of videos/DVDs, computers, general books, textbooks, electronics, and makeup. Conversely, students without a car at school were more likely to purchase these items online. Entertainment may be the one category that car owners were more likely to buy online because they have a car to be able to travel to the entertainment sites (e.g., concerts, sporting events) that may be located outside of the town where they attend school.

CONCLUSIONS

There are several interesting aspects of the results from this study. First, to really understand factors affecting Internet buying, it is important to analyze specific merchandise categories. It is apparent that the actual effects of the independent variables can be hidden when all merchandise categories are lumped together as one general equation—whether or not the respondent currently makes purchases on the Internet. This is also consistent with prior research on consumer purchasing (in other venues) in which the explanatory variables are different between total expenditures and individual expenditure categories, as well as among expenditure categories. Second, the variations that exist for specific independent variables among the merchandise categories are striking. These findings reinforce the need for Internet marketers to carefully segment their market based on factors such as age, gender, and income. Even within a relatively narrow demographic group (Generation Y college students), the need for refined target marketing seems imperative. Thus, using data mining and precision marketing would seem applicable to consumers within the Millenial generation. Third, as with purchases in non-Internet formats, variations in the effect of the explanatory variables seem to exist based on whether a category tends to be a necessity or a luxury. Although no specific economic testing was done to classify the merchandise categories as luxuries or necessities (because income was not measured as a continuous variable), commonsense interpretation indicates that

386

FAMILY AND CONSUMER SCIENCES RESEARCH JOURNAL

this differentiation appears to exist in the online environment. Further research could be conducted to examine such income effects. Another economic factor that could be explored is the role of prices. For example, do students from lower income families tend to buy textbooks and electronics online because Internet sites provide lower prices for these goods than traditional retail formats? Finally, it appears that the ability to identify a secure site increases the likelihood that a consumer will buy on the Internet. Even among the college student sample, which tends to be technologically savvy, 29% did not know how to identify a secure site.

IMPLICATIONS

These results have numerous implications for marketers, researchers, family and consumer sciences professionals, and policy makers. It is clear that these results suggest that marketers need to be very precise in defining their online target markets. Variations in the effect of socioeconomic and demographic characteristics, by merchandise categories, were demonstrated in this research. These results would suggest that “e-tailers” wanting to achieve the most efficient use of their marketing resources would pinpoint specific market segments for specific products, for example, younger college-age women for clothing and accessories, and older college-age men for computers. However, these results could also be used to help identify potential new market opportunities (i.e., market segments they don’t seem to be effectively reaching). Consumer researchers analyzing online buying behavior need to recognize the extent to which such variations exist and consider the viability of detailed categorical analyses. Further research could also build on this study by examining the actual dollar expenditures made on the Internet and defining income in such a way as to classify goods as luxuries or necessities. Family and consumer sciences professionals interested in consumer buying behavior may want to be informed about the issue of compulsive buying behavior. This is a factor that could be included in the financial education curriculum, as well as future studies on consumer buying behavior. Family and consumer sciences classes at the secondary and college levels could incorporate the DSCB scale into the curriculum as a way to create self-awareness about a student’s own shopping tendencies. Likewise, although irrational credit card use was not generally significant in this study, credit cards do affect shopping behavior as well as financial well-being. The irrational credit card usage scale could provide a starting point for classroom discussions concerning credit card use. Finally, the ability to identify a secure site seems to play an important role in Internet purchases. This variable had the same effect in all equations in which it was significant. If people can identify a secure site, they are more likely to make a purchase than if they cannot identify a secure site. E-tailers could benefit by educating their customers about how to identify a secure site and promoting the fact that their site is secure. From an educational perspective, family and consumer sciences educators involved with consumer and financial education might address Internet security as a component of buying behavior (including online purchasing). Stressing the importance of being able to identify a secure site is, in a sense, an extension of consumer protection education. Specifically, students should be taught the primary ways to identify a secure site. These include (a) whether the

Norum / INTERNET BUYING

387

site has a security certificate, (b) the presence of the lock icon on the screen, or (c) the presence of https (rather than http) in the URL. Providing such information is not intended to encourage Internet buying, per se, but to help students be informed consumers so they can protect themselves from fraud, identity theft, and so on. Policy makers may also be interested in this issue because identity theft is a growing consumer issue. As with any study, there were limitations that need to be considered when interpreting the results or planning future research studies. One limitation would be the sample. Although the sample size was sufficiently large, and all students enrolled in the campus had an opportunity to participate, greater diversity would be desirable. This could be achieved by extending the sample to universities in other geographic areas (domestic or international) and to noncollege students who are a part of Generation Y. Although Generation Y consumers are seen as technological innovators, their purchasing patterns may not reflect those of older consumers. Expanding the sample to other generations would allow intergenerational comparisons to be made. Research on Internet buying could be enhanced by obtaining more detailed information on actual expenditures for purchases online and on what affect purchasing online has on purchasing from traditional retail formats. REFERENCES Abdel-Ghany, M., & Silver, J. L. (1998). Economic and demographic determinants of Canadian households’ use of and spending on alcohol. Family and Consumer Sciences Research Journal, 27(1), 62-90. Dyer, B., Burnsed, K. A., & Dyer, C. L. (2006). Ethnicity and household expenditures: Furnishings, fashion and flux? Family and Consumer Sciences Research Journal, 35(2), 131-159. Edwards, E. A. (1993). Development of a new scale for measuring compulsive buying behavior. Financial Counseling and Planning, 4, 67-84. Faber, R. J., & O’Guinn, T. C. (1992). A clinical screener for compulsive buying. Journal of Consumer Research, 19(3), 459-469. Goldsmith, R. E., & Goldsmith, E. B. (2002). Buying apparel over the Internet. Journal of Product and Brand Management, 11(2/3), 89-102. Internet News. (2003, October 21). Stats. Retrieved January 23, 2007, from http://www.internetnews.com/ stats/article.php/3095681 Kim, Y., Kim, E. Y., & Kumar, S. (2003). Testing the behavioral intentions model of online shopping for clothing. Clothing and Textiles Research Journal, 21(1), 32-40. Kwak, H., Zinkhan, G. M., & Crask, M. R. (2003). Diagnostic Screener for Compulsive Buying: Applications to the USA and S. Korea. Journal of Consumer Affairs, 37(1), 161-169. Kwon, K., & Lee, J. (2003). Concerns about payment security on Internet purchases: A perspective on current on-line shoppers. Clothing and Textiles Research Journal, 21(4), 174-184. Lee, S., Lennon, S. J., & Rudd, N. A. (2000). Compulsive consumption tendencies among television shoppers. Family and Consumer Sciences Research Journal, 28(4), 463-488. Mitchell, S. (1998). American generations: Who they are, how they live, what they think (2nd ed.). Ithaca, NY: New Strategist. Norum, P. S. (2003). Apparel expenditure variation among U.S. households headed by baby boomers. Journal of the Textile Institute, 94, Part 2(1/2), 99-113. O’Guinn, T. C., & Faber, R. J. (1989). Compulsive buying: A phenomenological exploration. Journal of Consumer Research, 16(2), 147-157. Paulin, G. D. (2001). Variation in food purchases: A study of inter-ethnic and intra-ethnic group patterns involving the Hispanic community. Family and Consumer Sciences Research Journal, 29(4), 336-381. Phau, I., & Lo, C. (2004). Profiling fashion innovators: A study of self-concept, impulse buying and Internet purchase intent. Journal of Fashion Marketing and Management, 8(4), 399-411. Roberts, J. A. (1998). Compulsive buying among American college students: An investigation of its antecedents, consequences, and implications for public policy. Journal of Consumer Affairs, 32(2), 295-319.

388

FAMILY AND CONSUMER SCIENCES RESEARCH JOURNAL

Roberts, J. A., & Jones, E. (2001). Money attitudes, credit card use, and compulsive buying among American college students. Journal of Consumer Affairs, 35(2), 213-240. Rowley, M. L., & Okelberry, A. (2000). Shopping.com: Analysis of a trend. Journal of Family and Consumer Sciences, 92(1), 9-10. Shim, S., & Drake, M. F. (1990). Consumer intention to utilize electronic shopping. Journal of Direct Marketing, 4(3), 22-33. Shim, S., Eastlick, M. A., & Lotz, S. (2000). Assessing the impact of Internet shopping on store retailing among mall shoppers and Internet users. Journal of Shopping Center Research, 7(2), 7-43. Shoham, A., & Brencic, M. M. (2003). Compulsive buying behavior. Journal of Consumer Marketing, 20(2), 127-138. Spiggle, S., & Sewall, M. (1987). A choice sets model of retail selection. Journal of Marketing, 51(2), 97-111. Tacq, J. (1997). Multivariate analysis techniques in social science research. Thousand Oaks, CA: Sage. Taylor, S., & Cosenza, R. M. (2000). The impact of e-commerce on the merchandising of women’s clothing in traditional shopping centers/malls. Journal of Shopping Center Research, 7(2), 45-66. U.S. Department of Commerce. (2007). Quarterly retail e-commerce sales: 1st quarter 2007. U.S. Census Bureau News. Retrieved May 25, 2007, from http://www.census.gov/mrts/www/data/html/07/Q1 Varian, H. R. (1999). Intermediate microeconomics: A modern approach (5th ed.). New York: Norton. Watchravesringkan, K., & Shim, S. (2003). Information search and shopping intentions through Internet for apparel products. Clothing and Textiles Research Journal, 21(1), 1-7. Weiss, M. J. (2003). To be about to be. American Demographics, 25(7), 29-36. Wellner, A. S. (2000). Generational divide. American Demographics, 22(10), 52-58. Yin, S. (2003). Degree of challenge. American Demographics, 25(4), 20-22. Yurchisin, J., & Johnson, K.K.P. (2004). Compulsive buying behavior and its relationship to perceived social status associated with buying, materialism, self-esteem, and apparel-product involvement. Family and Consumer Sciences Research Journal, 32(3), 291-314.