A hard look at software
Do computer users rely two heavily on there spell-checking software? Andrea Everard thinks they doespecially since those programs often fail to recognize some common mistakes, including incorrect homonym use, such as “two” and “there” in the preceding sentence.
“People are far too trusting of computers,” Everard, assistant professor of accounting and management information systems in the Lerner College of Business and Economics, says. “A spell-checking program is certainly a good screening tool for writers, but it misses many errors.”
Everard, much of whose research focuses on human-computer interaction (HCI), conducted lab tests in which she and her colleagues asked volunteers to proofread an error-filled letter on a computer screen. One group of subjects was able to use the spell-checking software, while the other was told not to use it.
“We found that when people had the spell-checker on, they were much less likely to read the letter carefully,” Everard says. “In fact, native English speakers who relied on a spell-checking program didn’t do any better at proofreading than people who weren’t fluent in English.”
In addition, the researchers found that when a spell-checking program notes a possible error and offers an alternate spelling, people tend to accept the suggested revision even if it is incorrect or nonsensical, Everard says.
Everard began the spell-checking research at the University of Pittsburgh’s Katz School of Business, where she earned her doctoral degree in 2003, working with Prof. Dennis Galletta and fellow researchers Alexandra Durcikova and Brian Jones. Since joining the UD faculty later that year, she has extended her HCI research to examine such questions as how spelling and other types of errors on a web site influence consumers, as well as how consumers react to advertising on the web.
“When it comes to errors that people perceive on a web site, we’ve learned that it’s not, ‘Three strikes, and you’re out.’ It’s ‘One strike, and you’re out.’ People are very unforgiving if they encounter presentation flaws, and they have a very poor opinion of the site if they find, or just perceive, even one flaw,” Everard says. Those presentation flaws might be spelling mistakes, but they also might be design problems, such as missing photos or links that direct the user to sites that are under construction.
“What consumers think of the site then affects how likely they are to continue on it, to use it again or to buy things from it, so the research can provide useful information for online marketing,” Everard says.
The lesson for businesses using the web for marketing, she says, is to “test, test, test” every new site on as many focus groups and as many different types of consumers as possible before activating it. If people think there’s an error, even if there isn’t, they give the site a low rating, she says. “It’s all about perception, which is subjective,” she notes. “That’s even more reason to do a lot of testing before going live. Marketers should make their sites as perfect as they can.”
In addition to exploring how consumers are influenced by presentation flaws, Everard’s research looks at the effects of online advertising. Not surprisingly, she says, consumers are turned off by pop-up ads that interrupt their reading on a web site. But, she also has found that viewers find “pop-under” ads, which appear after a web site is closed and so do not interfere with the use of the site, just as bothersome.
“Apparently, people are still annoyed at having to do something extra [close the pop-under ad] before continuing on with their next task,” she says.
In general, Everard has found that banner ads, displayed across the top of the web site, are the least irritating to consumers. She also is looking at whether the subject matter of an ad affects consumers’ perception of it. Someone researching real-estate listings of homes for sale, for example, might find an ad about mortgage rates helpful rather than an intrusion. An ad on an unrelated topic, however, might be viewed with annoyance.
“I find that HCI is an area of information technology that’s fun to research,” Everard says. “It’s not highly technical, like programming, and so most people can relate it to their everyday lives.
“I want to have useful findings, rather than doing research just for the sake of research. This isn’t the kind of research that will save someone’s life, but I feel that I’m coming up with information that marketers and companies can actually use.”
As Everard continues her work with HCI, she also is researching cross-cultural issues in information technology. The daughter of an English father and Austrian mother, she was born in Africa and lived most of her life in cosmopolitan Montreal, where she spoke English at home and French in school. That background, she says, might be what has led to her interest in exploring different cultures.
“I started my research in cross-cultural issues by looking at the adoption of technology in South America, specifically in Uruguay and specifically examining the differences in technology in that country and the United States,” she says.
For the past two years, the focus of her international work has shifted to the Arab world.
Everard currently is working on her third article with Mohamed El Louadi, a colleague who now lives in his native Tunisia, on information technology in the Arab world. The two have collaborated on studies involving the digital divide, or the disparity in access to computer technology, between the West and the Arab world, as well as within the Arab world, where the wealthy Gulf states have vastly different technological resources and infrastructure than the region’s other nations.
She and El Louadi also are examining ways in which information technology (IT) can be brought into homes, enabling Arab womenwho often face cultural and religious restrictions on their movements and professional interactions with mento become employed and productive by telecommuting.
“All over the Arab world, men are learning IT skills and then leaving for jobs in Japan, the United States or western Europe,” she says. “So, just like in wartime, it’s going to fall to women to pick up the productivity and do those jobs at home.” In addition to benefiting their nation’s productivity, telecommuting from home could enable Arab women to make use of their skills and education without violating restrictions on such activities as driving, going out alone or speaking face to face with men, Everard says. She adds that if women are working with computers in their homes, their childrenparticularly their daughterswill become familiar with the technology, as well.
Arab women have not yet started telecommuting in great numbers, but the trend has started to take hold in Tunisia and is expected to expand, Everard says.
“If people say that women aren’t interested in technology, then we say: Bring the technology into the home, and they will become much more interested and also proficient,” she says.
Ann Manser, AS ’73, CHEP ’73