Job hirers show bias against regional accents, study suggests
If you make one-syllable words into two syllables, drop the "g" from words ending in "-ing," or stretch out the sound of the "a" in "car" and "park," will you encounter extra barriers while job searching?
Two UNT researchers say it's a strong possibility.
Patricia Cukor-Avila, assistant professor of English, and Dianne Markley, director of Cooperative Education and a graduate student in English, conducted the first known study on job hirers' judgments of American regional accents. The research was funded by grants from the Trice Foundation of Dallas and IBM.
As the director of Cooperative Education, Markley helps place students in internships and other positions while they are still working on their degrees. She says she decided to research regional accents for her master's thesis after learning about language attitude studies in Cukor-Avila's class.
"A study at Texas A&M University discovered that Texans and Southerners were more linguistically secure than those in other regions of the United States that is, we prefer our own accents," Markley says.
Around the same time, the Cooperative Education office received a phone call from an employer in New York who requested a student with a distinct Texas accent for an internship.
"The student would be calling Texas clients, and he said that Texans relate to and buy more from other Texans. His reason seemed legitimate, but something about screening candidates for a job based on their accents didn't seem right," Markley says. "I wondered if there was a way to measure if a regional accent would have an effect in the hiring process."
She points out that people's intelligence and other qualities are usually judged based on how they sound.
"I knew one woman a college relations director for a large, well-known Dallas-based corporation who had a Texas accent," Markley says. "She made a formal presentation at a professional meeting using a formal speaking style. Afterward, a man who had worked with her for years told her he never knew she was so smart. He had assumed she wasn't smart because of her accent."
With help from the UNT Center for Distributed Learning, Cukor-Avila and Markley created a CD-ROM program for human resource directors and others who hire new employees. The researchers recorded 10 males reading the same 45-second passage. Each speaker was from a different part of the United States Texas, Georgia, Louisiana, Alabama, North Carolina, Minnesota, California, Boston, Chicago and New Jersey. Each had an accent common to that region.
No speaker's age could be detected through his voice, Markley says, and she used only white male speakers, which limited the number of variables in the study. She adds that other factors, including gender, ethnicity and age, will be added in future studies if funding permits.
Fifty-six job hirers listened to the readers and made judgments about them based solely on how they sounded. The study participants were asked for their overall positive or negative impressions of each speaker, as well as whether the speaker seemed educated or uneducated, intelligent or not intelligent, energetic or lazy, uptight or laid back, outgoing or withdrawn, and assertive or docile.
They were also asked to classify each speaker as rough or refined, charming or irritating, and friendly or unfriendly, and to determine whether the speaker's background was urban or rural, cultured or earthy, and advantaged or disadvantaged. Using a U.S. map, the participants tried to guess each speaker's region. The participants then concluded whether each speaker would be competent or incompetent on the job and would fit into their companies' cultures.
Finally, the participants were presented with four job categories positions requiring a high level of public or customer contact; positions requiring a high level of technical expertise but little public or customer contact; positions involving extensive internal communications; and positions requiring little technical expertise and little or no customer contact. The participants determined which, if any, of the jobs the speakers would be suited for.
The speaker with a California accent was rated the most positively, followed by the speaker from Minnesota, the speaker from Boston and the speaker from Texas. The speakers from Louisiana, Georgia and New Jersey were rated the most negatively.
Cukor-Avila said less than 30 percent of the participants correctly identified the regions that the speakers came from. The accent of the speaker from New Jersey was the most recognized, whereas those of the speakers from California and Minnesota were the least recognized, she says.
Because few respondents placed an accent in the correct region, Cukor-Avila wonders if the hirers were reacting negatively to the actual accents or to the accents that they thought they heard.
"Generally, the less identifiable the accent was, the more highly the speaker was rated," she says. "This ties into the fact that in the United States, we tend to go for nondescript accents based on those of national broadcasters and actors. Advertising hasn't allowed regional sounds, because if a person selling a product sounds like he or she is from one part of the country, a person from another part of the country may not identify with him or her and may not buy the product."
The Texas speaker, she says, may have been highly rated because most of the job hirers participating in the study were from or had lived in Texas and only two of the 56 participants disliked their own accents enough to want to change them.
"Liking a Texas accent is usually tied into liking Texas in general. To be marked as a Texan through your speech is a positive thing for many people," Cukor-Avila says.
The North Carolina speaker may have been highly rated because respondents rated him as sounding cultured, she says.
In selecting job positions for the speakers, the study participants generally selected positions requiring a high level of public and customer contact for the speakers they rated the highest. They selected less visible positions for the lower-rated speakers, Markley says. In selecting speakers for positions requiring high-tech skills, the job hirers were less concerned about how friendly a person sounded than how educated he sounded, she says.
Markley says many of the study participants initially said they could not judge a person's intelligence, friendliness, economic background and other qualities based on 45 seconds of listening to the person.
"However, accent judgments hide behind judgments on education and other qualifications. An accent is a part of who you are. And unlike race or sex, it's not illegal to discriminate based on accent," she says.
She hopes that the survey in this study will eventually be made available to employers so that they can find out if job hirers discriminate against certain sounds.
"For instance, if someone had a very negative response to Georgia accents, a company would want to monitor that person's recruiting at places such as Georgia Tech," she says.
Other featured articles in this issue