The Effect of School Competitiveness on the Correlation Between Dogmatism and Learning Ability

Andrew Lee examines the ways in which the competitiveness of a school impacts the way in which the students learn information.


The current political environment is one of toxicity and irreconcilable disagreement. The heated disputes between the two main political groups, liberals, and conservatives, are caused by the dogmatic attitudes on both sides. Dogmatism is an aspect of a person’s personality encompassing their tendency to accept new beliefs that may contradict their own. It is evident how dogmatism affects the current debates between democrats and republicans, as neither side refuses to budge on their fundamental principles: supply-side versus demand-side economics, pro-choice versus pro-life, and gun control versus the right to bear arms. The bipartisan pervasiveness of dogmatism has been confirmed in a recent study done by Toner et al., which indicates that one side ceding to the other is an event that is unlikely to happen anywhere in the near future.1

Even though dogmatism manifests itself most evidently in the political sphere, it has a multitude of implications for education. Studies done by Costin and Ausubel and Tenzer confirm that dogmatic subjects have a harder time learning information that contradicts their beliefs.2 In the study done by Costin, research showed that more dogmatic students had a lower average score while attempting to learn information within a sociology class than within a psychology class, which was attributed to the fact that the sociology material included discussions of social issues while the psychology material did not, and as such there was no controversial material in the psychology class that would negatively affect a student’s learning ability.3 In the research done by Ausubel and Tenzer during the time period in which the Vietnam war was being fought, university students were assigned to either read a pro-Hanoi article or an anti-Hanoi article. The results showed that more dogmatic students who were already pro-Hanoi had a significantly harder time learning the anti- Hanoi material and vice versa.

Additionally, the differential in learning ability extends beyond a difference in learning information either in line with or contradicting one’s earlier beliefs. In a study done by Kemp, it was shown that more dogmatic students had harder times solving critical thinking problems,4 and in a study by Grippin and Ohnmacht, it was shown that more dogmatic subjects performed worse on a language learning exam.5 Lastly, in a study done by Maqsud and Khalique, results showed that more dogmatic students performed better in math classes.6 Critical thinking, language, and math all have nothing to do with pre-held beliefs about social issues, yet dogmatic students performed worse regardless, which provides evidence that dogmatism affects learning ability more generally.

The hypothesis that the researcher holds for why this effect occurs is because less dogmatic students are more likely to doubt their own beliefs, and thus are likely to check their own work. Although the researcher was unable to find research specifically correlating the tendency of students to double-check their work and higher grades, it is highly likely that students who take the time to review their work perform better.

No matter what the reason is for the existence of a differential in performance on tests unrelated to pre-existing beliefs between dogmatic and non-dogmatic subjects, the fact remains that they do perform better. Research in the past has attempted to confirm the validity of these results in a school environment by examining whether dogmatism had any correlation with performance within schools. A study by Shaver and Richards finds that students who have higher levels of dogmatism tended to perform worse in a methods-based social studies course,7 and a study by Stein and Furdon indicates that students who have higher levels of dogmatism tend to perform worse in social work graduate classes.8 The previously mentioned research done by Maqsud and Khalique was also conducted in an academic setting, as research was done at post-secondary institutions of education and universities in Bophuthatswana.

However, the school environments that these studies were conducted in were not highly competitive. The study done by Shaver and Richards was conducted at Utah State University, which is not a highly competitive university.9 The research done by Stein et al. was conducted in a graduate social work class, which is likely to be less competitive than the high school environment because high schoolers are competing against their peers to make them stand out to prestigious colleges and universities, while graduate students aren’t placed in such an environment. Maqsud and Khalique write in their research that students in Bophuthatswana are generally apprehensive of learning mathematics, indicating that there isn’t a highly competitive environment in mathematics education in Bophuthatswana.

This is important because of the nature of dogmatism. Dogmatism encompasses a person’s willingness to listen to new beliefs and their ability to accept those new beliefs. However, in a highly competitive environment, students are incentivized to accept any and all facts presented by their teacher, as a failure to do so will result in lower grades. The competitive mindset in a highly competitive school environment may thus be able to override the effect of dogmatism on student performance.

The researcher hypothesizes that the previous correlations between dogmatism and learning ability were due to a difference in confidence in oneself’s answers and methods. More dogmatic people were likely to believe themselves correct, which prevented rigorous examination of their own beliefs and thus dogmatic students fell behind because they never attempted to correct their answers, learning methods, or problem- solving methods.

However, in a highly competitive environment, students have more incentives to adopt new learning methods and question their own answers by virtue of the environment itself, since the social norm of high performance within these environments mean that students are likely more careful in their studying methods and are more cautious in their test answers. Thus, the researcher hypothesizes that there will be no difference in performance between dogmatic subjects and non-dogmatic subjects on the language learning exam due to the overriding effect of the competitive environment.

No current research has been done measuring the performance of students at different levels of dogmatism in a highly competitive school environment. Thus, this research will attempt to address this missing area of research by replicating previous methods of measuring the correlation between dogmatism and learning ability within a highly competitive high school environment.


The research was conducted at a high-SES High School in Austin, Texas. Although no formal measure for high school competitiveness exists, this school is widely considered a competitive high school, with an average SAT score more than 250 points above the state average, an average ACT composite score more than 6 points above the state average, an 82% Advanced Placement class participation rate, and was ranked as one of the top ten best public schools in Texas. Additionally, student testimony has confirmed the highly competitive environment, as students at this school have stated that stress and anxiety are mainly caused by the incentive for students to perform well and take difficult classes.

To measure dogmatism as related to student ability to learn information, two variables needed to be measured – dogmatism and the ability to learn information. Dogmatism was measured through the DOG scale, formulated by Altemeyer.10 This scale is a 20-question survey that allows for responses on a scale from -4 to 4 to questions such as “Anyone who is honestly and truly seeking the truth will end up believing what I believe” and measures dogmatism based on the responses. The scale is based off the Rokeach scale originally formulated by Milton Rokeach and was adjusted to measure dogmatism among students. Because it was specifically tailored to measure dogmatism among high school students, the DOG scale is likely to yield more accurate results than the Rokeach scale. Although using the original Rokeach scale would more closely emulate previous studies, the researcher determined that it would be beneficial to use the Altemeyer dogmatism scale for two reasons: one, the Altemeyer scale is much shorter, as it contains twenty questions instead of more than forty. This meant that the scale could be completed in a shorter amount of time, which decreased disincentives for participation. Two, the Altemeyer scale was more accurate for this population, since, as previously mentioned, the Altemeyer scale was tailored towards high school students – previous studies were conducted with subjects sampled from undergraduate and graduate populations for which the Rokeach scale may have been more appropriate, while this research was conducted in a high school setting.

The second variable that required measurement was a student’s ability to learn information. To minimize the influence of prior knowledge of a topic, the researcher administered a foreign language test that measures a student’s ability to learn common Dutch words. Other subjects, such as mathematics tests or trivia, would allow for previous knowledge to influence student performance. One limitation of using language as a method to measure learning ability is that previous research done by Grey et al. found that bilingual people find it easier to pick up languages, although separate research done by Abu-Rabia and Sanitsky found that the influence of bilingualism is not large.11 12 Dutch was chosen because English and Dutch are very similar due to their shared nature as West Germanic languages. The US Department of State specifically states that Dutch is one of the easiest languages to learn due to its similarity to English, with data compiled from over seventy years of teaching diplomats.13 This means that both languages share similar word structures, including usage of the same letters, which facilitated easier learning of the language. That allowed the researcher to reduce the amount of time required for studying, which lessens disincentives to attend the study. To minimize the influence of English language knowledge on knowledge of Dutch, the researcher chose Dutch words that are not similar to their English counterparts. To minimize the influence of knowledge of Dutch or other languages similar to Dutch, a question was asked after the test about whether the participant had any previous knowledge of Dutch, German, Afrikaans, Swedish, Norwegian, or Flemish, since these are all West Germanic languages similar to Dutch and thus students who answer “yes” to this question had their data excluded to prevent skew of the data.

Another reason for the usage of a language test is that previous research done by Grippin and Ohnmacht included a nearly identical language exam, with the only difference being that Grippin and Ohnmacht administered a Russian language test rather than a Dutch language test. The researcher chose Dutch instead of Russian because, as indicated by the US Department of State, Dutch is simply easier to learn, which allowed the researcher to reduce the amount of time required for research and thus reduce disincentives to participate in the research. Dutch is categorized by the US Department of State as a Category I Language, indicating that it is similar to English and thus easy to learn, while Russian is categorized as a Category III Language, indicating that it has significant differences from English and thus is more difficult to learn.

Participants were given a list of 25 Dutch words and their English translations and were told to attempt to learn all the information presented to them. The time limit for this learning period was ten minutes, which was imposed for two reasons. First, a time limit reduced the total amount of time required for the research, which incentivized more people to participate in the research. Second, it has been shown that less dogmatic students spend more time studying,14 which may have also held true within this study – thus, the time limit was imposed to eliminate the variable of the amount of time spent studying that may be affected by dogmatism. The list of information follows the format of:

Allemaal Altogether

Weg Road

(The entire list of words and dogmatism survey will be included in the appendix.)

Next, participants were instructed to log on to a school-provided electronic device and fill out a Google form that tested them over the previously learned language information. This Google form was set to “quiz” mode, which allowed for easy scoring of the answers. Participants were not allowed to see their results right after they finish the quiz because doing so would 1] slow down the research, as students would begin to discuss the results with each other if they were made available to them and 2] doing so could skew the data – if a student knew that they had a low score, they could attribute that to having a higher level of dogmatism and could hold that mindset (that they are a highly dogmatic subject) while taking the dogmatism test, which could skew results. The data were then entered into a separate spreadsheet that tracked correlations between dogmatism and language learning ability.

After participants are finished taking the language test, they will take the DOG survey. This was administered through a Google form and results were gathered via a Google sheets spreadsheet linked to the Google form. The format of the google forms survey was as such: the form posed a statement such as “Anyone who is honestly and truly seeking the truth will end up believing what I believe” and respondents were allowed to pick an answer from nine different choices to indicate how accurately that statement aligns with their beliefs: 1] “Extremely inaccurate”, 2] “Very inaccurate”, 3] “Somewhat inaccurate”, 4] “Slightly inaccurate”, 5] “Neither accurate nor inaccurate”, 6] “Slightly accurate”, 7] “Somewhat accurate”, 8] “Very accurate”, 9] “Extremely accurate”. These scores were recorded within the previously mentioned spreadsheet as text and Google sheets functions were used to translate each answer into a score corresponding to the question. For example, if the question posed above was asked, those that answered “Slightly inaccurate” would have a 4 assigned to their score for that question, and those that answered “Very accurate” would be assigned a score of 8, as those who believe that the statement is representative of their true beliefs are likely more dogmatic. Scores were tallied up within the spreadsheet using another formula. The data were then entered into the previously mentioned spreadsheet that tracks the correlations between the two variables.

To randomly sample participants from the population of the high school at which the research was conducted, the researcher used a spreadsheet containing a list of all students within the high school after obtaining permission to do so from the school. This list also contained the students’ emails and grade level. Excel’s random function (=RAND) was then used to select 250 random numbers between 1 and 3008, and the rows of the spreadsheet corresponding to the randomly selected numbers were retrieved, and students who were randomly selected for the study were sent an email containing a Google form asking whether they would be able to participate and a parental permission form. Out of the first sample of 250, 9 respondents indicated that they would be able to participate, with four people showing up on the day of the study. Since a sample size of four is not large enough to determine whether the statistics were normally distributed, another sample was taken. Additionally, at this point in the research, the parental permission form requirement was removed as the researcher stopped collecting names of subjects and thus a permission form was no longer necessary. This was done to remove disincentives from participating in the research, as more requirements to participate mean that fewer people will participate.

Another four hundred people were selected using Excel’s random function. Using Excel’s filter function overlaps between the first sample of 250 and the second sample of 400 were removed from the second sample, and emails were sent out to the emails from the second sample that did not overlap. Out of the second sample of 400, 12 people indicated that they were able to participate, and six people came on the day of the study. Since the data set only numbered ten people after both samples, the researcher decided to conduct one last sample. One thousand people were randomly sampled using Excel’s random function, with overlaps with the previous 650 people eliminated from the sample using previously mentioned functions. Additionally, the researcher created even more incentive to participate by holding the study during both fourth and fifth-period classes.

The high school at which this research was done operates on an eight-period system, meaning that the school day is divided into eight equal periods, with one class each period. Each student also has either fourth or fifth period bracketed out for lunch. Because the researcher had previously conducted the research only during the fourth period, the sample size was effectively halved, as those that had lunch during the fifth period were unable to participate. By conducting the research during both periods, the researcher increased the number of people that were able to participate.

Ultimately, the large sample size and the fact that research was conducted during both periods meant that twenty people showed up on the date of the study. After this point, the researcher decided to stop data collection and begin data analysis.


To determine whether the data were significant, the researcher used a Linear Regression T-test

conducted using the data collected from the survey, with the data of subjects who answered that they had previous knowledge of Germanic languages similar to Dutch removed. In total, the number of subjects with usable data was 26.

It was found that the correlation between the two sets of data had nearly no correlation with an r-value of -0.06749 and a p-value significantly above the alpha. This means that dogmatism likely has zero effect on language learning ability within the context of this study, and any possible correlation between the two sets of data was most likely due to random chance.

These data confirm the researcher’s hypothesis, as they show that, within a highly competitive environment, there is no difference in performance between low and high dogmatic subjects. We can state with some confidence that these data show the strength of the overriding effect of a competitive environment, as there were nearly a few differences between this research and the research done by Grippin and Ohnmacht. Both this research and the research done by Grippin and Ohnmacht shared the usage of a foreign language test and measured dogmatism through the Rokeach scale or a scale similar to the Rokeach scale, although Grippin and Ohnmacht administered a matched-pairs language test while this research used a free-response style language test. One major difference between the two is that the research done by Grippin and Ohnmacht did not impose a time limit. The absence of a time limit may be significant, as the research by Robbins and Rogers indicates that more dogmatic students study for a longer amount of time. However, their research also indicates that, although more dogmatic students study for a longer amount of time, this difference in study time is not associated with a difference in performance, meaning that the usage of a time limit in this research should not make a large difference.

An additional conclusion that can be drawn is that less dogmatic subjects do not perform better in school because they have increased mental capacity or naturally better learning ability; rather, the differential exists as a result of different attitudes that dogmatic and non-dogmatic subjects take towards learning. The only adjusted variable between this variable and previous research was the context in which it was done, which affects student attitudes. If it is true that dogmatic students are simply better at learning, then the less dogmatic students in this study should have performed better in the language test. Because there is nearly zero correlation between the two measured variables in this research, the hypothesis that less dogmatic students are naturally better at learning is false.

Discussion and Limitations

The results of this research indicate that a highly competitive environment overwhelms the effect of dogmatism on student learning ability. This has a couple of implications for future research.

The first implication is that research that attempts to measure correlation between a certain attitude and student performance within an academic setting will likely need to conduct their research in a less competitive environment such as a college or a less competitive high school to prevent the effect of competitiveness from hiding the effect of student attitudes on student performance.

The results of this research could also be used as a justification for repeating certain studies in a different environment. For example, if a previous study conducted in a highly competitive environment concluded that there was no association between the attitude measured within that research and student performance, this research indicates that repeating the research in a different environment may yield different results.

However, there are a few limitations within the study that prevent certainty in the results. The first limitation is the existence of a non-response bias. Although random sampling was used, a very small proportion of those who were sampled showed up to participate. It is likely that only those students who were interested in being a part of academic research participated within the study, and those students that are interested in academic research are likely to be students that are more competitive than less. However, although normally a non-response bias would be a limitation, the non-response bias in this research is a benefit. If it is true that only highly motivated and competitive students participated in the study, then the sample more accurately represents a highly competitive environment.

The second limitation of this research is that it does not completely emulate previous research like the research done by Grippin and Ohnmacht. As previously mentioned, this study has a few key differences. First, this research uses the Altemeyer DOG scale instead of the older Rokeach dogmatism scale that Grippin and Ohnmacht used. This change did not majorly affect the results, as, according to Altemeyer, the only difference between the two is a single question. The second difference between this research and previous research is that, in this study, Dutch was used instead of Russian. This should not make a large difference, as the researcher finds no evidence that the process of learning Dutch is massively different from the process of learning Russian. The third difference is the method in which participants were given the information they were to learn. According to Grippin and Ohnmacht, the participants were given a packet of information in a lesson-like format where example sentences containing Russian and English translations were used, and their research did not mention a time limit. This research imposed a time limit and information was given in a format with only the English word and its Russian translation, which is dissimilar to the lesson-like format used by Grippin and Ohmacht. However, there should not be a large difference in results. If the lesson-like format administered by Grippin and Ohnmacht changed learning ability, it should have changed learning ability by the same amount across the board, as making the information easier to learn would have made the information easier to learn for all subjects.

Because all of the differences are not large, it can still be stated with a reasonable degree of confidence that this research closely emulates a previous study in the same area of research although the emulation is not entirely accurate.

The third limitation of this research is that there are variables that could not be controlled between this research and previous research, meaning that school competitiveness is not the only variable that changed between the studies and thus the change from a correlation between dogmatism and learning ability to the lack of a correlation could be due to different variables. Factors that changed between the studies include time period and school demographics. Most of the previous research in the area of dogmatism and learning ability was conducted in the 1970s, with the only recently conducted study done by Maqsud and Khalique.

As for demographics, the area where this research was conducted was a high-SES majority Caucasian high school. Previous research was conducted at universities in America and Bophuthatswana, where students have less disposable income, with more than a third of college students struggling with basic needs such as food and housing.15 As such, there exists a clear resource disparity between where research was previously conducted and current research.

No previous research has been done on whether these variables significantly change dogmatism’s effect on learning. Thus, it is possible that the variables of time period and demographics could each individually be the actual reason that dogmatism has no effect on student learning ability. However, this is unlikely as the researcher finds no evidence that either of these variables could have an effect on the correlation between dogmatism and learning ability.

Because the limitations mentioned are unlikely to significantly impact the results of the research, the researcher has determined that the results are more likely than not to be correct. Thus, the implications of this research for future research in the area of student attitudes and performance seem clear. The competitive

educational environment has more significant effects on student performance than dogmatism. Continued research in this area may be able to determine whether other variables such as demographics and time period also affect the correlation between dogmatism and learning ability and whether this effect is generalizable to other student attitudes and environments, i.e. whether other environmental aspects are sufficient to override the effects of other attitudes on student performance.

Appendix A – Language List

Vandaag Today

Morgen Tomorrow

Gisteren Yesterday

Kunnen Can

Gebruiken Use

Klein Small

Mooi Beautiful

Lelijk Ugly

Moeilijk Difficult

Makkelijk Easy

Slecht Bad

Dichtbij Near

Heerlijk Delicious

Wijn wine

Rundvlees Beef

Lichaam Body



Verpleegster Nurse

Werknemer Employee

Lerares Teacher

Verkoper Salesman

Alles everything

Weg Road

Misschien Maybe

Mensen People

Allemaal Altogether

Tegen Against

Dingen Things

Genoeg Enough

Graad Gladly

Appendix B – Dogmatism Survey

1.       Anyone who is honestly and truly seeking the truth will end up believing what I believe (M = 3.92, SD = 2.53).

  • There are so many things we have not discovered yet, nobody should be absolutely certain their beliefs are right (M = 4.02, SD = 2.51). R
  • The things I believe in are so completely true, I could never doubt them (M = 4.44, SD = 2.38).
  • I have never discovered a system of beliefs that explains everything to my satisfaction (M = 3.86, SD = 2.52). R
  • It is best to be open to all possibilities and ready to reevaluate all your beliefs (M = 3.39, SD = 2.16). R
  • My opinions are right and will stand the test of time (M = 4.63, SD = 2.25).
  • Flexibility is a real virtue in thinking, since you may well be wrong (M = 3.12, SD = 1.85). R
  • My opinions and beliefs fit together perfectly to make a crystal-clear “picture” of things (M = 4.45, SD = 2.21)

  • There are no discoveries or facts that could possibly make me change my mind about the things that matter most in life (M = 4.57, SD = 2.52).
  • I am a long way from reaching final conclusions about the central issues in life (M = 3.75, SD = 2.21). R

  1. The person who is absolutely certain she has the truth will probably never find it (M = 4.39, SD = 2.23). R
  2. I am absolutely certain that my ideas about the fundamental issues in life are correct rect (M = 4.77, SD = 2.23).
  3. The people who disagree with me may well turn out to be right (M = 4.00, SD = 2.08). R
  4. I am so sure 1 am right about the important things in life, there is no evidence that could convince me otherwise (M = 4.04, SD = 2.29).
  5. If you are “open-minded” about the most important things in life, you will probably reach the wrong conclusions (M = 3.08, SD = 1.79).

  1. Twenty years from now, some of my opinions about the important things in life will probably have changed (M = 3.41, SD = 2.15). R
  2. “Flexibility in thinking” is another name for being “wishy-washy” (M = 3.34, SD = 1.99).
  3. No one knows all the essential truths about the central issues in life (M = 3.23, SD = 2.24). R
  4. Someday I will probably realize my present ideas about the BIG issues are wrong. (M = 5.24, SD = 2.10). R
  5. People who disagree with me are just plain wrong and often evil as well (M = 2.33, SD = 1.66).

Works Cited

1 Toner, K., Leary, M. R., Asher, M. W., & Jongman-Sereno, K. P. (2013). Feeling Superior Is a Bipartisan Issue: Extremity (Not Direction) of Political Views Predicts Perceived Belief Superiority. Psychological Science, 24(12), 2454-2462.

2 Ausubel, D., & Tenzer, A. (1970). Components of and Neutralizing Factors in the Effects of Closed-Mindedness on the Learning of Controversial Material. American Educational Research Journal,7(2), 267-273. Retrieved from

3 Costin, F. (1965). Dogmatism and learning: A follow-up of contradictory findings. The Journal of Educational Research, 59(4), 186-188. Retrieved from JSTOR database. (Accession No. 27531689)

4 Kemp, C. G. (1963). Improvement of critical thinking in relation to open-closed belief systems. The Journal of Experimental Education, 31(3), 321-323. Retrieved from JSTOR database.

5 Grippin, P. C., & Ohnmacht, F. W. (1977). Field independence and dogmatism as mediators of performance on a programmed learning task with and without strong prompts. The Journal of Experimental Education, 45(4), 13-15. Retrieved from JSTOR database.

6 Maqsud, M., & Khalique, C. M. (1991). Relationships of some socio-personal factors to mathematics achievement of secondary school and university students in Bophuthatswana. Educational Studies in Mathematics, 22(4). Retrieved from JSTOR database.

7 Shaver, J. P., & Richards, H. E. (1971). Open-closed mindedness and an inquiry-oriented social studies methods course. The Journal of Educational Research, 65(2), 85-93. Retrieved from JSTOR database.

8 Stein, S., Linn, M. W., & Furdon, J. (1974). Predicting social work student performance. Journal of Education for Social Work, 10(3), 85-92. Retrieved from JSTOR database.

9  Niche. “Utah State University.” Retrieved from

10 Altemeyer, Bob. Dogmatic Behavior Among Students: Testing a New Measure of Dogmatism, The Journal of Social Psychology, (2002) 142:6, 713-721, DOI: 10.1080/00224540209603931

11 Grey, Sarah et al. Bilingual and monolingual adults learning an additional language: ERPs reveal differences in syntactic processing. Bilingualism: Language and Cognition, Volume 21, Issue 5. November 2018 , pp. 970-994.

12 Abu-Rabia, Salim and Sanitsky, Ekaterina. (2011, February 1). Bilinguals find it easier to learn a third language. ScienceDaily. Retrieved April 7, 2019 from

13 U.S. Department of State. “FSI’s Experience with Language Learning.” U.S. Department of State.

14 Robbins, G. E., & Rogers, D. E. (1975). Dogmatism and Study Time in High School Students: Better to be Wrong than Long. The Journal of Education Research, 69(3), 120-121. Retrieved from JSTOR database.

15 Hess, Abigail. “New study finds that 36% of college students don’t have enough to eat.” April 6th 2018. CNBC. Retrieved from enough-to-eat.html

Leave a Reply

Your email address will not be published. Required fields are marked *