The Effects of Accents on Perceiving Speakers' Truthfulness

The Effects of Accents on Perceiving Speakers’ Truthfulness
by
Takiyah R. Walcott
Submitted to the Board of Study in Psychology
School of Natural and Social Sciences
in partial fulfillment of the requirements
for the degree of Bachelor of Arts
Purchase College
State University of New York
May 2016
Sponsor: Ager Gondra
Second Reader: Karen Singer-Freeman

2
ACCENTS AND TRUTHFULNESS
Abstract
The current work investigates if the perception of foreign accents influences people’s judgments
about the reliability of speakers. Twenty-one participants were played 20 recorded statements
(10 by each speaker) that were read by a European speaker or a Latin American speaker. Both
speakers had Spanish accents and subjects were asked to rate how truthful the statements were
on a scale of 1-5. Upon completion, subjects were asked to write if they noticed anything unusual
about the statements that might have impacted their decisions about the truthfulness of the
statements. We hypothesized that participants would find the Spain speaker to be a more reliable
source than the Dominican speaker. However, we failed to find difference between how many
statements spoken by the European were true and how many statements spoken by the
Dominican were judged to be true.

3
The Effects of Accents on Perceiving Speakers’ Truthfulness
An accent is a unique form of pronunciation in a language often associated with
nationality and/or a speaker’s dialect (Edwards, 1997 as cited by Carlson & McHenry, 2006). It
is a personal characteristic that is sometimes associated with a positive stereotype or negative
stereotype. The stereotype itself can either enhance or belittle the way a person is viewed. For
instance, French English accents and British English accents are often associated with
sophistication, intelligence, and charm. Latin American accents on the other hand are unfairly
criticized, associated with low status, and limited education, success, and intelligence (Brennan,
1997 as cited by Fuertes, 1999).
Though there were many choices to choose from of European accents and Latin
American accents, this experiment intends to focus on accents from Spain and the Dominican
Republic. A European accent in comparison to Latin American accents is a field that has not
gotten much attention. The hypothesis of this study is that when speaking English, people with
Spanish accents from Spain will be considered a more trustworthy/reliable source than people
with Latin American accents (Spanish from the Dominican Republic).
Intelligence can be defined in many different ways but it is most often associated
with an individual’s thoughts, communication skills, understanding, creativity, logic, and
knowledge (Fuertas et al., 2002). It is quite interesting that a distinctive mode of pronunciation as
a result of speaking a foreign language can cause assumptions about one’s communication skills,
understanding, and logic. Research findings from psycholinguistics studies suggest that people
sometimes use speech accents to make inferences about the background and personality traits of

4
an individual (Fuertes et al., 2002). These accents can sometimes affect the way listeners view an
individual in terms of attractiveness, social class, and compatibility. Researchers began to come
up with experiments and theories that could possibly explain this phenomenon.
The Accent Prestige theory by (Giles & Powesland, 1975) was developed to
explain why specific accents were associated with intelligence and why others were not. Giles
and Powesland tested their theory and found that speakers in the United Kingdom whose accents
had “first class” accents were often viewed with higher prestige regarding “status” and
“solidarity” than speakers who were not from the UK (Fuertes et al., 2002). Other studies were
done to help understand the assumption of how an accent could possibly help determine the traits
and characteristics of an individual. A study in the United States examined the effects of middle
class White, middle class Black, and “ghetto” Black accents on the evaluations of African
Americans and White listeners inferences about kindness and honesty (Fuertes et al., 2002).
Fuertes et al. that the middle class White and middle class Black accents were not rated
differently on personality traits such as kindness and honesty. However, Ghetto Black accents
received significantly lower ratings of kindness and honesty when rated by White listeners. The
African American listeners did not rate any of the accented groups differently based on their
accents.
Accents seem to have an effect on many things ranging from social class and
personality traits to even compatibility between people. This realization caused the American
Psychological Association to start coming up with better ways to improve the relationships
formed within counseling and/or research. The method they were coming up with was to help
counselors become more open-minded when interacting with their patients. In 1993, the

5
American Psychological Association’s Board of Minority Affairs came up with a set of
guidelines to encourage mental health professionals to consider and understand cultural
differences such as language differences and interaction styles to help develop better
communication (Fuertes et al., 2002). Interestingly enough, maybe it was not the counselors that
needed to accept/respect the cultural differences, but the patients themselves.
Fuertes (1999) conducted two studies examining how different races perceive
Hispanic counselors. The first study tested Asian Americans and African Americans and the
second study tested Euro-Americans. The counselors were White Hispanic, Mestizo Hispanic,
and Black Hispanic. Fuertes (1999) found it important to test Asian Americans and African
Americans because studies from multicultural counseling suggest that minorities sometimes have
a preference for people of the same race. The Hispanic counselors used in this experiment either
had a thick Spanish accent when speaking English or no Spanish accent at all. The participants’
initial perceptions of the counselors were examined using the Universal-Diverse Orientation
(UDO) by Miville et al. (Fuertes, 1999). The UDO measures attitudes about tolerance, openness,
and respect for differences such as gender, race, physical abilities, and sexual orientation
(Miville, 1992 as cited by Fuertes, 1999). Three outcome measures were used to evaluate initial
perceptions of the counselors: the Counseling Rating-Form Sheet (Corrigan & Schmidt, 1983),
the Working Alliance Inventory-Short (Horvath & Greenberg, 1986, 1989; Tracey & Kokotovic,
1989), and a Likert-type Willingness Scale (Fuertes, 1999). The Counseling Rating-Form Sheet
measured the participants’ ratings of the counselor on expertness, attractiveness, and
trustworthiness. The Working Alliance Inventory-Short measured participants’ expectations and
anticipation of building a counselor-client relationship and the Willingness scale measured how
willing the participants were to work with the counselor long term (Fuertes, 1999). Fuertes

6
(1999) found several interesting results. Participants that scored high on the UDO expressed that
they would enjoy building a therapy relationship with their counselors and were also very willing
to work with the counselors long-term. Fuertes (1999) did not find that the counselors’ race and
accent affected the way Asian Americans and African Americans viewed them. In Fuertes’
(1999) second study that tested Euro-Americans initial perception of Hispanic counselors, the
results were quite different.
Fuertes, Jairo, Gelso, and Charles (2000) conducted a study to see how White
Hispanic, Mestizo Hispanic, and Black Hispanic counselors were perceived by Euro-Americans
(Whites) It was important to examine Euro-Americans in this study because previous research
has shown that race is important to Euro-Americans in shaping perceptions of themselves as well
as others (Fuertes et al., 2000). The Hispanic counselors either had thick Spanish accents or no
Spanish accent at all. It was important to distinguish between the different complexions of the
Hispanic counselors because there was a possibility of a preference. This study utilized the same
measures that were described in Fuertes et al., (1999). Fuertes et al. (2000) found that
participants were more willing to work long-term with the Hispanic counselors that did not have
a Spanish accent than Hispanic counselors that did have an accent. Participants that scored low in
UDO found the counselors with no accent to be more attractive, trustworthy, and professional
than the counselors with an accent. An interesting result was that some of the participants’
preferred to work with the Black Hispanic counselor more than the White Hispanic and Mestizo
Hispanic counselor. Fuertes et al. (2000) explained that the participants’ who had this preference
scored high in UDO. People who score high in UDO are said to find interest in understanding
and learning about the differences of other races and Fuertes et al. (2000) believed that this
perspective could explain why some Euro-Americans preferred the Black Hispanic counselor.

7
These two studies had interesting results regarding perceptions of accented speakers in a
professional setting so it is only right to look at the possible effects of an accent affecting
employability as a whole.
Carlson and McHenry (2006) were interested in the effects an accent or dialect
could have on employability. A dialect is a bit different than an accent. A dialect resembles a
variety of elements of the language that is spoken by the majority but still differs in certain ways
such as syntax and semantics. In this article, Carlson and McHenry describe three major
categories of ethnic accents that are different from Standard American English: Spanish
Influenced English, Asian Influenced English, and African American Vernacular English
(AAVE). For this particular study, the Asian and Spanish influenced speakers were considered to
be “accented speakers” due to phonological features and the AAVE were considered as dialectal
speakers due to morphological differences. Carlson and McHenry hypothesized that
bidialectilism is a good characteristic for speakers to have (the ability to change two different
speaking styles based on social expectations). For instance, when speaking to an elder or giving a
presentation, using the SAE dialect (Standard American English) may be appropriate and when
talking to your peers, using AAVE dialect may be appropriate. For the experiment, three female
bidialectal speakers applied for the same job. The females were given scripts to read in which
they were required to switch their accents and dialect at different times. The actors recorded
several versions of each statement that they made and these recordings were later played for 60
adults that were working in the field of human resource management. Carlson and McHenry
found that speakers of Spanish-influenced English AAVE were rated higher than Asian-
influenced speakers of English in understanding because their speech style was almost similar to
SAE. However, the Asian-influenced speaker was rated as the second highest for employability

8
despite the participants rating her low on comprehension skills which could possibly indicate that
it is not only the accent of the speaker but the stereotype about them that influences
employability. The lowest rating of employability was given to the AAVE speaker. Carlson and
Henry hypothesize that this may have resulted from negative associations the participants had
about African Americans. If accents influence employability, might they also have an effect on
one’s comfort in the work place?
Previous research has found that discrimination can cause a great amount of stress
and affect an employee’s work ability. A study conducted by the U.S General Accounting Office
(1990) reported 10% of the employees surveyed had been discriminated against because of their
accents (Ryan, Hewstone, & Giles, 1984 as cited by Wated & Sanchez, 2006). Wated,
Guillermo, & Sanchez, Juan I., 2006) conducted an experiment to examine if accents result in
greater stress at work. Researchers measured organizational commitment, job satisfaction, and
work tension among Hispanic workers that have a Spanish accent when speaking English. Wated
and Sanchez hypothesized discrimination would be negatively associated with employee
outcomes. Hispanic students that spoke English with a Spanish accent and were employed full-
time answered questions assessing their organizational commitment, accent-based perceived
discrimination, job satisfaction, role ambiguity, and work tension. Wated and Sanchez found that
discrimination can indeed be a powerful work stressor for employees that are minorities. As for
their first hypothesis, Wated & Sanchez (2006) did not find evidence to support the claim that
accent-based discrimination would predict organizational commitment. Researchers did find that
coping strategies such as self-efficacy, perceived control, and group identity did not seem to help
with accent-based perceived discrimination (Wated & Sanchez, 2006). With several articles

9
about Hispanics and the perception of them based on their accents, it is finally time to look at a
study regarding European accents.
Jeahyeon (2011) was interested in seeing how an instructor’s accent could influence
a student’s achievement and how a students’ accent preference could affect their learning. This
study had three components: survey, instruction, and assessment. Before starting the experiment,
participants were asked to complete a survey assessing their knowledge of statistics, accent
perceptions, and multimedia learning (internet based pictures). The participants were shown a
short tutorial video explaining how to use SPSS (a software program) with an unaccented, a
mildly or heavily European accented, or a mildly or heavily Asian accented voice giving the
instructions. Jeahyeon found that there was not much of a difference between the students’
achievement level whether they heard a native voice or an accented voice. However, students
that did not like Asian accents showed lower performance than those who didn’t mind Asian
accents. European accents were not related to performance. The native voices were higher than
both the European and Asian accented ratings but the European accented rating was much higher
than the Asian accented rating.
Most of the experiments mentioned above examined the effects of African-
American, Asian, and Latin American accents on listeners’ judgments. The previous research
shows that there is indeed a bias towards European accented speakers in comparison to Latin
American accented speakers. The purpose of this senior project is to explore a particular field
that has not gotten much attention: people’s perception on trustworthiness and reliability in a
speaker from Spain and Speaker from DR.

10
Method
Participants
Twenty-one people participated in this experiment (13 females, 8 males). All of the
participants were born and raised in the United States of America. Subjects were asked to
participate voluntarily and did not receive any compensation. Participants’ ages ranged from 18-
26 and 10 were White, 6 African Americans, 3 Latin Americans, and 2 Asians.
Design
This experiment used a within-subjects design where the independent variable was
Accents (European and Latin American) and the dependent variable was truthfulness.
Materials
Two men (one from Spain, one from the Dominican Republic) were recorded reading 19
statements in English. They were both males with a high school degree and some college
education. The speaker born in Spain was 31 years old and the speaker born in the Dominican
Republic was 23 years old. Due to experimental error, one statement was read by a male speaker
with an American accent. In order to have the same number of statements read with each accent,
the statement read with an American accent and one other statement were excluded. The
recordings were anywhere from 4-21 seconds long. The statements were divided into two
groups. Group 1 included the European speaker reading 9 statements and the Dominican reading
the other 9 students. Group 2 included the same statements read by the speaker with the other
accent. The statements were ordered so that the speakers were alternated.

11
Procedure
Participants were tested both individually and in groups of 2 people. The recordings
were anywhere from 4-21 seconds long. Participants were played each recorded statement twice
and then asked to circle whether they thought the statement was true (yes or no). Immediately
after, they were asked to rate on a scale of 1-5 how true they believed the statement was with 1
being “not true” and 5 being “true.” After rating the statements, participants’ were given another
sheet of paper that said “Did you notice anything about the recordings? If so, did it impact your
responses in any way?” The participants’ originally heard 20 statements (8 or 9 from each
speaker depending on which group they were in and one from an American speaker) but 2
statements were excluded. The Americans’ statement was excluded because it was not relevant
to my hypothesis. The second statement was excluded in order to maintain the same number of
statements in each accent. We selected a statement which the majority of subjects believed was
false. Upon completion of the task, participants were verbally debriefed and told they would
receive an email regarding the results if they were interested.
Result
I was interested in seeing if people would find a European accent to be a more reliable
source of information than a Latin American accent. Participants judged the Dominican’s
statements to be true an average of 4.67 times (SD = 1.77). They judged the European’s
statements to be true an average of 5.05 times (SD = 1.86). A paired samples t-test was utilized
to assess if people would find the statements of the European accented speaker to be true more
frequently than the statements of the Latin American accented speaker. Results revealed no
significant difference between how many statements spoken by the European were judged to be

12
true and how many statements spoken by the Dominican were judged to be true t(20) = -.84, p =
.41.
When asked to rate how likely a statement was to be true using a 1(not truthful) to
5 (truthful) point Likert scale, participants rated the Dominican speaker’s statements’
truthfulness as averaging 2.95 (SD = .75) and the European speaker’s statements’ truthfulness as
averaging 3.06 (SD = .69).
A paired samples t-test was utilized to assess whether people found the European
accented speaker to be more likely to be true than the Dominican accented speaker. We failed to
find a significant difference t(20) = -.67, p = .52.
Discussion
The purpose of this senior project was to explore whether people’s perception of the
truthfulness of statements would be influenced by the type of Spanish accent a speaker had.
Although the Dominican-speaker’s statements were found to be true less frequently that the
European-speaker’s statements. Contrary to my hypothesis, this difference was not large enough
to be reliable. Similarly, when asked to rate how likely it was that the statements were true,
participants rated the likelihood that the Dominican speaker’s statements were true as lower than
the likelihood that the European speaker’s statements were true. However, once again the
difference was not large enough to be reliable. In both cases, the speaker from Spain was trusted
more than the speaker from DR which helps support my hypothesis even though the results were
not significant. Only 21 participants were included in the current study. Given that the responses

13
differed in the predicted pattern, a larger sample might have yielded significant results. For
future research, using a bigger sample size may prove helpful.
After completing the experiment, participants asked whether they noticed anything
particular about the statements. Though the participants’ responses varied, a common response
was that they could not understand some of the speakers’ statements. One participant specifically
said “I could not understand the Asian speaker.” Though each statement was played twice,
participants’ still asked for certain statements to be replayed (primarily statements made by the
Dominican speaker). The Dominican speaker did have a thicker accent than the speaker from
Spain so clarity could have possibly been an issue.
Another potential flaw of the current study was our choice of statements. Many
of our statements were historic. Participants who knew that some statements were true would not
be influenced by the speaker’s accent. Conversely, some statements sounded impossible.
Participants were likely to say these statements were untrue regardless of the speaker’s accent as
well. A way to avoid this problem would be to include autobiographical statements. The
participants would be unable to use their prior knowledge to assess the truthfulness of these
statements.
Using more accents from both Europe and Latin America (Spanish and French
from Spain and France vs. Spanish and French from DR and Haiti) could also be helpful.
Assuring that the speakers’ accents are on the same level regarding clarity would be a good idea
as well. This experiment showed that there could possibly be a difference in people’s judgments
of accents on truthfulness/ reliability. For future research, this can be retested and previous
research has shown that people judge attractiveness based off accents as well so that is also
something that can be tested for.

14
References
Ahn, J. (2011). The effect of accents on cognitive load and achievement: The relationship
between students' accent perception and accented voice instructions in students'
achievement Available from PsycINFO. (871544367; 2011-99090-050). Retrieved from
http://search.proquest.com/docview/871544367?accountid=14171
Carlson, H. K., & McHenry, M. A. (2006). Effect of accent and dialect on employability.
Journal of Employment Counseling, 43(2), 70-83. doi:10.1002/j.2161-1920.2006.tb00008.x
Fuertes, J. N. (1999). Asian Americans' and African Americans' initial perceptions of Hispanic
counselors. Journal of Multicultural Counseling and Development, 27(3), 122. Retrieved
from http://search.proquest.com/docview/235996509?accountid=14171
Fuertes, J. N., & Gelso, C. J. (2000). Hispanic counselors' race and accent and Euro Americans'
universal-diverse orientation: A study of initial perceptions. Cultural Diversity and Ethnic
Minority Psychology, 6(2), 211-219. doi:http://dx.doi.org/10.1037/1099-9809.6.2.211
Fuertes, J. N., Potere, J. C., & Ramirez, K. Y. (2002). Effects of speech accents on interpersonal
evaluations: Implications for counseling practice and research. Cultural Diversity and
Ethnic Minority Psychology, 8(4), 346-356. doi:http://dx.doi.org/10.1037/1099-
9809.8.4.347
Giles, H., & Sassoon, C. (1983). The effect of speaker's accent, social class background and
message style on british listeners' social judgements. Language & Communication, 3(3),
305-313. doi:http://dx.doi.org/10.1016/0271-5309(83)90006-X

15
Wated, G., & Sanchez, J. I. (2006). The role of accent as a work stressor on attitudinal and
health-related work outcomes. International Journal of Stress Management, 13(3), 329-350.
doi:http://dx.doi.org/10.1037/1072-5245.13.3.329

16
Appendix
20 statements played.
1. It is possible for twins to be born up to 5 days apart.
2. There are more Chinese restaurants in the US than McDonalds, Burger King, KFC, and
Wendy’s combined.
3. In Mexico, artists can pay taxes with their art work
4. A woman sold a chicken nugget on eBay for $8,100 because it looked like George
Washington.
5. More salt is used for de-icing roads than for human consumption
6. The oldest condoms ever found date back to the 1640s and they were made from animal
and fish intestines.
7. Maine is the closest US state to Africa
8. The Romans used to clean and whiten their teeth with urine
9. Nintendo was originally a trading car company.
10. The US president must pay for their own meals.
11. George Washington died on December 14, 1799
12. There is only one country between North Korea and Norway
13. There are more stars in space than there are grains of sand on every beach in the world.
14. Camels can hold a grudge and wait for the opportunity to take revenge
15. Scotland’s national animal is the unicorn
16. Oxford University is older than the Aztec Empire
17. Vin Diesel broke into a theatre in New York with his brother when he was 7 to vandalize
it. Instead of calling the police, the theater’s artistic director offered him a role in an
upcoming show which started Vin’s acting career
18. Everyone has a unique tongue print, just like finger prints.
19. Armadillos almost always give birth to quadruplets
20. For every human on Earth there are 1.6 million ants

The Effects of Accents on Perceiving Speakers' Truthfulness

Recommended

Recommended

More Related Content

Viewers also liked

Viewers also liked (16)

Similar to The Effects of Accents on Perceiving Speakers' Truthfulness

Similar to The Effects of Accents on Perceiving Speakers' Truthfulness (20)

The Effects of Accents on Perceiving Speakers' Truthfulness