Empirical evidence of deep learning in learning communities

Kara Smith (Massey College of Business, Belmont University, Nashville, Tennessee, USA)
Robin Lovgren (College of Sciences and Mathematics, Belmont University, Nashville, Tennessee, USA)

Journal of Applied Research in Higher Education

ISSN: 2050-7003

Publication date: 2 July 2018



The purpose of this paper is to investigate whether learning communities (LCs), defined as a cohort of students jointly enrolled in two distinct courses, increase “deep learning” in either or both courses. This study focuses on the impact of learning communities in quantitative courses.


The hypothesis is tested using a unique data set including individual student performance and characteristics collected from students enrolled in an LC of Principles of Microeconomics and Elementary Statistics. The sample also includes students enrolled in each course separately which allows for testing between groups. The final exam in each course contained questions designed specifically to test deep learning. The design facilitates the use of multivariate regression analysis to examine the correlation between learning in communities and deep learning, holding constant other possible elements of student success.


Despite perceptions among the sample student population that learning increases in both courses as a result of the LC format, the empirical evidence does not reveal any statistically significant increase in deep learning as a result of learning in community. However, the sample is more introverted than the average college student which may meaningfully impact the results.

Research limitations/implications

There are a number of important motivations for implementing an LC program that are not measured here. These include an increased sense of community among students, breadth (rather than depth) of knowledge, and awareness of the interconnectedness of learning across disciplines. However, to the extent that university instructors are motivated to ensure learning in their own discipline, this resource-intensive strategy may not be the most suitable approach in quantitative courses.


Learning communities continue to be a popular pedagogical technique and curriculum requirement, particularly at teaching-focused universities. This research offers an empirical approach to measuring one aspect of their value which is typically left to conceptual or qualitative study.



Smith, K. and Lovgren, R. (2018), "Empirical evidence of deep learning in learning communities", Journal of Applied Research in Higher Education, Vol. 10 No. 3, pp. 311-321. https://doi.org/10.1108/JARHE-11-2017-0141

Download as .RIS



Emerald Publishing Limited

Copyright © 2018, Emerald Publishing Limited


College instructors are faced with the perennial task of engaging students with course content in a way that maximizes their learning for the long run. Ken Bain, author of What the Best College Teachers Do (2011), suggests that the best college teachers work to promote deep learning. In other words, a rigorous education should encourage students to understand rather than merely remember. Stephen Buckles and John Siegfried (2006) study deep learning in economics, specifically, and define deep learning as a student’s ability to use reason to find the correct answer to a question even when there are multiple steps and those steps are undefined. Ultimately, students should understand topics, rather than simply remember bits and pieces. This is particularly true in quantitative fields where students are prone to fall back on memorization.

One possible way to increase engagement and deep learning in quantitative courses is to offer two courses as a learning community (LC). LCs have become more popular along with an increased interest in collaborative learning (Tinto, 2003; Schroder, 2010). Improved student outcomes have been well-documented for students learning in community (Smith and MacGregor, 2009; Buch and Spaulding, 2008). Positive outcomes include increased student engagement (Buch and Barron, 2011), increased student metacognition and learning through a combination of LCs and self-assessment (Siegesmund, 2016), improved student perceptions of learning gains (Wingert et al., 2014), higher student retention (Love, 2012; Barrie, 2016), and increased retention rates of transfer students (Lord et al., 2012). LCs also mitigate the disintegration of higher education into individual, disconnected courses (Huber et al., 2005; Love, 2012). Kuh et al. (2008) recommend LCs as a high impact activity and as one of the options available in higher education to enhance student engagement and increase student success. LCs are even shown to increase deep learning (Chapman et al., 2005; MacGregor et al., 2000; Mahoney and Schamber, 2011). We propose to test this result empirically in an LC linking two highly quantitative courses.

Although an LC may take on many forms, the LC investigated here is what Lenning and Ebbers (1999) define as a “curricular learning community.” This is in contrast to the three other types of student LCs: classroom, residential, and student type LCs. Curricular LCs were originally designed to accommodate an increasing number of students commuting to campus and not engaging to a large degree outside of the classroom. Zhao and Kuh (2004) find that for commuter students, curricular LCs are related to greater academic and social involvement. They suggest that more research is needed to determine whether some types of LCs are more educationally effective than others for certain groups of students and for which outcomes. This research adds to that specific body of knowledge.

The curricular LC studied here is comprised of a cohort of 30 students at a mid-sized private Christian university in the south who register for two courses (Principles of Microeconomics and Elementary Statistics) in back-to-back class blocks and complete joint assignments throughout the semester. They do not continue into other classes with the same cohort and their living arrangements are not affected by the LCs.

The instructors participate in one another’s LC course approximately once every three or four weeks. The statistics instructor shares examples of economics problems and how to solve them statistically, while the economics instructor consistently references statistics as a way to back up and test economic principles. The joint project is done in stages through the semester. Students gather data and do statistical analysis to define a relationship between two variables, then explain the economic implications of the results. A rubric is created jointly by both instructors and the grading is done together. In addition to continual contact through e-mail, the instructors meet regularly through the semester to discuss process improvements, to deal with student work/performance issues, to work on other non-related interdepartmental projects, and to socialize, ensuring a united appearance to the students. This research proposes to empirically test the potential for LCs to enhance deep learning in highly quantitative courses. To do so, we generate a unique set of data on 178 students enrolled in Principles of Microeconomics and Elementary Statistics, some of whom are enrolled in the courses as a part of an LC. The final exam of each course contained questions specifically designed to test for deep learning. Regression analysis is used to estimate the causal effect of enrolling in the courses as a part of an LC on deep learning, controlling for other individual student characteristics, such as GPA, that may be correlated with learning outcomes.

In addition, due to the expectation of increased social interaction between students in LCs, this research adds to prior work by measuring each student’s level of introversion. Mull (2006) gives an overview of literature describing introverts and extroverts. Introverts tend to be self-focused and shy, quiet, retiring, introspective, serious, studious, reserved, passive, pessimistic, tend to plan ahead, seldom lose their temper, have a few intimate friends, and value ethical standards (Eysenck and Eysenck, 1975, 1994). Cattell (1946) claims introverts are preoccupied with inner ideas and emotions while Good (1959) describes introverts as tending to withdraw socially. Importantly, introversion can be considered a negative trait in college students (Henjum, 1982). Henjum explains that even though introverts can have positive traits such as being analytical, self-sufficient, and hard-working, they are pressured to be participative, outgoing, and social in modern college settings which may make them question if they are normal.

This paper extends important research on a widely used approach to student learning to understand its potential to improve deep learning in quantitative subjects, Principles of Microeconomics and Elementary Statistics, and by including a measure of student introversion. Furthermore, the research is unique because of an experimental design which includes treatment and control groups. This approach facilitates measurement of the marginal change in deep learning as a result of studying Microeconomics and Elementary Statistics jointly in the LC environment.

Interdisciplinary learning to promote deep learning

In a review of 70 assessment studies of LCs, MacGregor et al. (2000) find, promisingly, that students in LCs do exhibit deeper learning. They also find that LCs boost academic outcomes, particularly for disadvantaged students, and that students and faculty both benefit socially and have a greater sense of community. Mahoney and Schamber (2011) investigate the value of LCs as a mechanism to promote deep learning using a qualitative approach. Analyzing the text of student speeches in an LC linking a first-year general education seminar and a public speaking course, the authors conclude that LCs can, indeed, be a fertile ground for deep and integrative learning.

A primary and intentional objective of LCs is to enhance social connections between students (Jaffee et al., 2008; Arensdorf and Naylor-Tincknell, 2016). In fact, cultivating a sense of community is often a primary motivator for the implementation of an LC program (Brownell and Swaner, 2010). However, that sense of community depends on individual student characteristics in addition to the environment into which a student enters. DeNeui (2003) discusses the role introversion plays on students’ psychological sense of community their first year in college. According to DeNeui, extroverted students are primed to find community more quickly than less extroverted students. Furthermore, Borg and Shapiro (1996) note that introverted students performed better with independent learning as compared with group learning in economics courses. Given the inherent emphasis on student interactions and collaborative learning within LCs, it is possible that student introversion meaningfully impacts whether learning in community enhances deep learning. The inclusion of introversion as a covariate helps us to understand the generalizability of our findings to other experimental settings.

Research setting and design

In this experiment, students are enrolled in an LC comprised of Principles of Microeconomics and Elementary Statistics. Each course is required for students earning Bachelor of Business Administration degrees and is an option for most undergraduate students in fulfillment of the university’s general education requirements. As a result, hundreds of students sign up for these courses each semester. Second semester freshmen signing up for these courses have the option to take them jointly as an LC. (All students are required to take part in an LC but not this particular pairing of courses.) Students are strongly encouraged to complete an LC course pair during the spring semester of their freshman year. Our study population, therefore, consists of three groups: student enrolled in Principle of Microeconomics only, students enrolled in Elementary Statistics only, and students enrolled in both courses as a part of an LC.

It is worth noting that the university in this study does not explicitly state deep learning as a goal of the LC pedagogy. Rather, the focus is on the cross-disciplinary nature of learning. All three groups of classes (the LCs, Elementary Statistics only, and Principles of Microeconomics only) enjoy the benefits of active learning as recommended by Lenning and Ebbers (1999). It is common practice in all Elementary Statistics classes for the students to practice the collaborative learning strategy think-pair-share, to work problems on worksheets using technology, and to work in small groups at the board. Microeconomics active learning practices include group problem solving, economic games, and current event analysis.

In the LC, the same cohort of 30 students are enrolled in the two courses in back-to-back class blocks. In addition to double the class time with the same classmates, the students complete a substantial project which brings together knowledge from the two disciplines and counts for 15 percent of the course grades in each class. The interdisciplinary assignment requires them to work in small groups to collect original or publicly available data to test one of Mankiw’s Ten Principles of Economics (Mankiw, 2015). There are five intermediate assignments over the course of the semester which lead to the final project presentation. This assignment challenges students to use knowledge from each course in a cross-disciplinary way throughout the semester.

Students enrolled in the non-LC version of each course cover the same material and take the same tests. They do not, however, complete the interdisciplinary assignment described above. Instead, in Principles of Microeconomics, students take one additional mid semester test. In Elementary Statistics, students complete a statistical project on an arbitrarily chosen data set.

While the University’s primary goal is to help students see that knowledge exists and learning can occur both within and between disciplines, the added layer of complexity presented by cross-disciplinary work may push students to a deeper understanding of each discipline. Student comments on course evaluations reflect a perception of increased learning in each course that is attributable to the format, consistent with Wingert et al. (2014). For instance, one student comments that “…the relationship between the two, especially micro and stats, is really valuable to understand…I’m grateful for taking this course as a linked course because I learned so much more than I would have by taking them individually.” This paper’s objective is to test this archetypical observation empirically.

During the Spring 2015 semester, each author taught four sections of the same course in their fields, Principles of Microeconomics and Elementary Statistics, respectively. Two sections in each topic were linked into two separate LCs. The other four sections (two in each topic) were populated with students not enrolled in the other instructor’s course. (There were three students who, entirely by chance, enrolled in both courses but not as a part of an LC.) Therefore, the sample includes 59 students enrolled in both courses as a part of an LC, 69 students enrolled only in Principles of Microeconomics and 54 students enrolled only in Elementary Statistics.

Two surveys were administered to each student.

The first was a basic demographic survey administered electronically by email using Qualtrics. Data collected include student’s GPA, academic classification (freshman, sophomore, etc.), major, expected course grade, gender, race, and age.

The second survey was the McCroskey introversion scale. This 18-question instrument was embedded in the Qualtrics survey. Scores on this instrument must be between 12 and 36 where a score greater than 28 indicates high introversion and less than 20 indicates low introversion (Richmond and McCroskey, 1997).

The authors rely on McCroskey’s Introversion Scale due to its brevity relative to more extensive personality tests (such as Meyers-Briggs) and its narrow focus on measuring introversion. This specific tool excludes questions tied to apprehension in communication and thus can be used to test for introversion independent of apprehension in communication. The questions McCroskey uses are from the work of Eysenck (1970, 1971) with the communication apprehension questions omitted. Eysenck’s scales on Extroversion-Introversion were shown to have statistically significant correlation with the Extroversion-Introversion scales of the Myers Briggs tests thus showing convergent validity at the self-report questionnaire level. The Personal Report of Communication Apprehension (PRCA-24) (McCroskey, 1993) is a longer version of the introversion test including questions on communication apprehension and is considered highly reliable (α regularly>0.90). It is considered to have very high predictive validity.

McCroskey’s work has been validated and accepted in the communication studies body of knowledge (McCroskey et al., 2014; Bragg, 2017). McQueen (2012) uses the PRCA-24 with LCs, finding that students in an LC have considerably higher communication competence scores at the end of the semester. When compared with the PRCA-24, the Introversion Scale’s α reliability estimates were measured above 0.80. Furthermore, McCroskey’s Introversion Scale has been applied to measure introversion in university student populations, giving us a benchmark to compare our sample pool (Richmond and McCroskey, 1997).

At the end of the semester the cumulative final exams in each course included questions designed specifically to test for deep learning. The questions for Microeconomics come directly from Buckles and Siegfried (2006). The questions for Elementary Statistics were designed in a similar fashion to the Microeconomics questions. Both sets of test questions were in the format of assertion-reason questions (ARQs), a sophisticated form of multiple choice questions that promote higher-order thinking from the students. Williams (2006) asserts that ARQ multiple choice questions can serve as substitutes for, and can predict performance on, other types of questions, such as short answer and essay questions. Huntley et al. (2009) find that well-designed multiple choice questions can effectively evaluate complex thinking and learning in undergraduate mathematics courses.

As an example, the question below, from Buckles and Siegfried (2006), relates to economic externalities in markets, but that is not explicitly stated. In order to answer the question correctly, the student must recognize the topic and correctly use a supply and demand model to analyze the effect of an externality:

If the production of a product sold in an otherwise competitive market generates waste that does not affect the producer but does injure or annoy other people:

a. price and quantity are both too high to generate an efficient allocation of resource.

b. price and quantity are both too low to generate an efficient allocation of resources.

c. price is too high and the quantity is too low to generate an efficient allocation of resources.

d. price is too low and the quantity is too high to generate an efficient allocation of resources.

The experimental design is unique because of the presence of both experimental and control groups and quantitative data on individual student characteristics and learning. The sample includes students enrolled in each course separately (the control group) and as a part of an LC (the treatment group). Furthermore, all sections of each topic were taught by the same instructor (one for Microeconomics, one for Statistics) ensuring maximum consistency in content delivery. The final exams in all four sections of each course, from which outcome data are collected, were identical in each section of the two courses. This experimental design facilitates the estimation of the marginal impact that learning in a community can have on the desirable but elusive goal of deep learning.

The authors note that learning is a complex process and we do not aim to model that process in its entirety. Other works such as Mamerow (2014) aim to unravel the “why” questions behind enhanced learning and LCs. However, this is beyond the scope of the work presented here.

Summary statistics and student characteristics

Table I lists summary statistics of the sample student population broken down by course enrollment. LC students (i.e. students enrolled in both courses as a part of an LC) tend to be younger and report higher GPAs on a four-point scale. Both findings are consistent with the University’s recommendation that students enroll in the required LC during the Spring semester of the freshman year. This is also evident in the academic classification[1] of each sample group. Early in their academic career, students are just beginning to take more difficult courses which are likely to reduce their overall GPA.

Students in our sample are somewhat introverted. While the average college student has an introversion score of 19 (Richmond and McKrosky, 1997), the average student in this sample scores as somewhat introverted with a score between 23 and 24. This difference is particularly important as we analyze the impact of a pedagogy which functions as a mechanism to enhance students’ interaction and sense of community.


Table II summarizes student performance on the final exams in each class and on 11 questions designed specifically to test for deep learning as described above. Although students enrolled in the LC of Elementary Statistics and Principles of Microeconomics did slightly better on the final exams than students enrolled in one of the courses independently, the difference is not statistically significant. There is not a consistent pattern on the difference between specific deep learning questions except that none of the differences are statistically significant.

In order to dig deeper into this question, regression analysis is used to test the causal relationship between the percent of deep learning questions answered correctly in these two courses and multiple factors which may affect grades. The following equation is tested:

% DLQ i = β 1 LCC i + β 2 IN i + β 3 LCC i × IN i + β 4 X i + μ i
where the dependent variable is the percent of deep learning questions answered correctly by the student. The variable LCC indicates whether the student is enrolled in the course as a part of an LC. The student’s introversion is indicated by the variable IN. These two variables are also interacted to indicate whether a higher level of introversion is correlated with a marginally different impact of learning in community. The vector Xi includes the student characteristics summarized in Table I above. Results from this regression are detailed in Table III.

In microeconomics, performance on questions measuring deep learning is mostly driven by student aptitude, as measured by GPA. Older students also appear to exhibit more deep learning. For both courses, there is no statistically significant evidence that deep learning is different for students enrolled in the course as a part of an LC or the effect of the LC is correlated with student introversion.

Prior research (such as Smith and MacGregor, 2009; Buch and Spaulding, 2008) suggests that students perform better in courses that are part of an LC. In order to test our data against this finding, we repeat the previous regression using the student’s cumulative final exam grade as the dependent variable. The following equation is tested:

E x a m G r a d e i = β 1 L C C i + β 2 I N i + β 3 L C C i × I N i + β 4 X i + μ i
where the dependent variable is the student’s final exam grade in either Principles of Microeconomics or Elementary Statistics. These results are presented in Table IV.

There is no clear evidence that being a part of an LC increases student performance in either course in a statistically significant way. Any measurable effect of the LC format is swamped by student aptitude, as measured by GPA and self-reported expected course grade. Although the results are not reported here, the same regression was repeated using each deep learning question as a dependent variable. The results were fundamentally unchanged.

While these results run counter to the limited prior research on LCs and deep learning, there are number of possible explanations. First, this experiment uses a relatively small sample which may limit our statistical power. Second, the students in this sample are more introverted than the average college student. As Henjum (1982) explains, the enhanced social environment of an LC may not be a good fit for an introverted learner. Finally, and perhaps most importantly, this experiment tested an LC of two highly quantitative subjects, in contrast to earlier research by Mahoney and Schamber (2011). The findings presented here may suggest that LCs are best suited to enhancing deep learning in qualitative courses.

Conclusions and implications for practice

Ample prior research demonstrates the many benefits of LCs. These benefits range from increased student engagement to higher student retention and more. Deep learning in relation to LCs has been studied under certain conditions. A review of LCs by MacGregor et al. (2000) finds that students in LCs exhibit deeper and more integrated learning. Mahoney and Schamber (2011) review a curricular LC composed of a first year general education course and a public speaking course and find an increase in deep learning.

This research is designed to add to the body of knowledge on deep learning in LCs by looking specifically at a curricular LC involving two quantitative courses and testing for deep learning empirically. The research presented is a test of whether enrolling in Principles of Microeconomics and Elementary Statistics as a part of an LC increases deep learning in either of the two fields relative to independent enrollment. Furthermore, this experiment introduced a measure of introversion to investigate whether a student’s introversion is correlated with an increase in deep learning. Ultimately, the findings indicate that the curricular LC environment, as defined by Lenning and Ebbers (1999) and as carried out in this study, is not a statistically significant factor, on the margin, for ensuring deep learning in these two fields.

These findings do not contradict any current research findings, but rather bring into question the specific conditions for deep learning in an LC. The specifics for this experiment include a curricular LC with instructors visiting each other’s classes every three to four weeks, two quantitative courses, and students who are more introverted than traditional college students. Any one of these factors, or the combination, may have led to the lack of statistical significance in the results.

It is also important to note that deep learning in each field distinctly is not a learning goal of the LC as defined by the University. Instead, the primary goal is to encourage students to recognize and appreciate the interconnectedness of knowledge across disciplines. This is an important and untested objective within the experimental design. Fortunately, success on this front is evidenced in the qualitative comments from students. Therefore, these results should not be read as a disincentive for universities to explore the use of LCs to encourage students to be healthy, well-rounded learners, increase retention, or to cultivate a sense of belonging.

However, college instructors are strongly incentivized to ensure learning in their discipline first and foremost. Higher-order objectives such as student retention and interdisciplinary thinking are often secondary. Active learning strategies, and LCs in particular, are resource intensive and require far more time investment by faculty than a course that is taught independently. Therefore, faculty and administrative officers must carefully consider the strategic priorities at both the university and classroom level. This research suggests that LCs may be more suited to advancing broad, university-level objectives (at least in quantitative courses) than the primary classroom objective of disciplinary learning.

Student summary statistics

Variable Learning community Microeconomics only Statistics only
Current GPA 3.46 3.29 3.30
Academic Classification 1.25 2.32 1.70
Expected course grade 3.51 3.39 3.44
McCroskey introversion 23.0 24.0 23.5
% female 49.2 48.5 37.0
Age 18.8 20.0 19.4
N 59 69 54

Dependent variable summary statistics

Mean (score or % students answering correctly)
Deep learning metric LC ECON only STATS only Difference of means p-value
Microeconomics final exam 72.179 (1.567) 71.371 (1.687) 0.808 (2.330) 0.729
Statistics final exam 80.932 (1.391) 76.796 (2.294) 4.136 (2.632) 0.119
Econ DLQ1 0.695 (0.060) 0.681 (0.056) 0.014 (0.083) 0.868
Econ DLQ2 0.864 (0.045) 0.942 (0.028) −0.078 (0.052) 0.135
Econ DLQ3 0.271 (0.058) 0.261 (0.053) 0.010 (0.079) 0.896
Econ DLQ4 0.525 (0.066) 0.449 (0.060) 0.076 (0.089) 0.394
Econ DLQ5 0.525 (0.066) 0.492 (0.061) 0.033 (0.089) 0.752
Econ DLQ6 0.746 (0.057) 0.739 (0.053) 0.007 (0.078) 0.932
Stats DLQ1 0.780 (0.054) 0.868 (0.047) −0.088 (0.073) 0.227
Stats DLQ2 0.390 (0.064) 0.340 (0.064) 0.050 (0.092) 0.586
Stats DLQ3 0.610 (0.064) 0.679 (0.065) −0.069 (0.091) 0.451
Stats DLQ4 0.508 (0.066) 0.623 (0.067) −0.114 (0.094) 0.228
Stats DLQ5 0.746 (0.057) 0.831 (0.052) −0.084 (0.078) 0.281
N 59 69 54

Notes: Standard errors in parenthesis. *Indicates a difference of means at the 5 percent level

Marginal impact of learning communities on deep learning

Dependent Variable: % of deep learning questions correct
(1) (2)
Variables Microeconomics Statistics
LCC =1 is student enrolled in LC 0.241 (0.179) −0.243 (0.243)
Introversion per McCroskey 0.002 (0.006) 0.001 (0.008)
Introversion × LCC −0.010 (0.007) 0.006 (0.010)
Current GPA (4.0 Scale) 0.131*** (0.046) 0.069 (0.066)
Expected Course Grade (4.0 Scale) 0.077* (0.040) 0.088 (0.063)
Female =1 if student is female 0.058 (0.036) 0.069 (0.051)
Age 0.030* (0.017) 0.005 (0.019)
Observations 119 102
R2 0.189 0.122

Notes: Standard errors in parentheses. Constant term not reported. *p<0.1; **p<0.05; ***p<0.01

Marginal impact of learning communities on final exams

Cumulative final exam score
(1) (2)
Variables EconFinalExam StatFinalExam
LCC −2.205 (8.228) 6.662 (12.121)
IntroLCC −0.034 (0.339) −0.136 (0.512)
Current GPA 12.656*** (2.110) 5.754* (3.312)
Expected_Grade_4scale 6.305*** (1.833) 11.430*** (3.072)
Introversion −0.043 (0.253) 0.064 (0.397)
Female1 1.305 (1.666) −0.065 (2.586)
Age 0.549 (0.774) 1.483 (0.963)
Observations 119 103
R2 0.471 0.303

Notes: Standard errors in parentheses. Constant term not reported; *p<0.1; **p<0.05; ***p<0.01



This variable is used to indicate whether each students is a freshman (variable is coded as 1), sophomore (2), junior (3), or senior (4).


Arensdorf, J. and Naylor-Tincknell, J. (2016), “Beyond the traditional retention data: a qualitative study of the social benefits of living learning communities”, Learning Communities: Research & Practice, Vol. 4 No. 1, p. 4.

Bain, K. (2011), What the Best College Teachers Do, Harvard University Press, Cambridge, MA.

Barrie, L.F. (2016), First-year Students: Investigating the Impact of Participating in a Mathematics Learning Community, master’s thesis, Rowan University, Glassboro, NJ.

Borg, M.O. and Shapiro, S.L. (1996), “Personality type and student performance in principles of economics”, The Journal of Economic Education, Vol. 27 No. 1, pp. 3-25.

Bragg, J. (2017), Communication Apprehension Among Community College Students: A Phenomenology, doctoral dissertation, East Tennessee State University, Johnson City, TN.

Brownell, J.E. and Swaner, L.E. (2010), Five High-impact Practices: Research on Learning Outcomes, Completion and Quality, Association of American Colleges and Universities, Washington, DC.

Buch, K. and Barron, K.E. (2011), “Increasing student engagement through curricular-based learning communities”, Promoting Student Engagement, pp. 63-69.

Buch, K. and Spaulding, S. (2008), “Using program assessment to ‘prove and improve’ a discipline-based learning community”, Journal of Learning Communities Research, Vol. 3 No. 2, pp. 35-46.

Buckles, S. and Siegfried, J.J. (2006), “Using multiple-choice questions to evaluate in-depth learning of economics”, The Journal of Economic Education, Vol. 37 No. 1, pp. 48-57.

Cattell, R.B. (1946), Description and Measurement of Personality, World Book Company, Oxford.

Chapman, C., Ramondt, L. and Smiley, G. (2005), “Strong community, deep learning: exploring the link”, Innovations in Education and Teaching International, Vol. 42 No. 3, pp. 217-230.

Deneui, D.L. (2003), “An investigation of first-year college student’s psychological sense of community on campus”, College Student Journal, Vol. 37 No. 2, pp. 224-235.

Eysenck, H.J. (1970), Readings in Extraversion-introversion: Volume I, Wiley Interscience, New York, NY.

Eysenck, H.J. (1971), Readings in Extraversion-introversion: Volume II, Wiley Interscience, New York, NY.

Eysenck, H.J. and Eysenck, S.B.G. (1975), Manual of the Eysenck Personality Questionnaire (Junior and Adult), Hodder and Stoughton, London.

Eysenck, H.J. and Eysenck, S.B.G. (1994), Manual of the Eysenck Personality Questionnaire: (EPQ-R Adult), EdITS/Educational and Industrial Testing Service, San Diego, CA.

Good, C.V. (1959), Dictionary of Education, Mc-GrawHill Book Company, New York, NY.

Henjum, A. (1982), “Introversion: a misunderstood ‘individual difference’ among students”, Education, Vol. 103 No. 1, pp. 39-42.

Huber, M.T., Hutchings, P. and Gale, R. (2005), “Integrative learning for liberal education”, Peer Review, Vol. 7 Nos 3/4, pp. 4-7.

Huntley, B., Engelbrecht, J. and Harding, A. (2009), “Can multiple choice questions be successfully used as an assessment format in undergraduate mathematics?”, Pythagoras, Vol. 69, pp. 3-16.

Jaffee, D., Carle, A., Phillips, R. and Paltoo, L. (2008), “Intended and unintended consequences of first-year learning communities: an initial investigation”, Journal of the First-Year Experience & Students in Transition, Vol. 20 No. 1, pp. 53-70.

Kuh, G.D. and Schneider, C.G., Association of American Colleges Universities (2008), High-impact Educational Practices: What They are, Who has Access to Them, and Why They Matter, Association of American Colleges and Universities, Washington, DC.

Lenning, O.T. and Ebbers, L.H. (1999), “The powerful potential of learning communities: improving education for the future”, ASHE-ERIC Higher Education Report, Vol. 26, ERIC Clearinghouse on Higher Education, One Dupont Circle, NW, Washington, DC.

Lord, V., Coston, C., Blowers, A., Davis, B. and Johannes, K. (2012), “The impact of a discipline-based learning community on transfer students: a multi-dimensional pilot study”, Journal of The First-Year Experience & Students in Transition, Vol. 24 No. 2, pp. 63-84.

Love, A.G. (2012), “The growth and current state of learning communities in higher education”, New Directions for Teaching and Learning, Vol. 2012 No. 132, pp. 5-18.

McCroskey, J.C. (1993), An Introduction to Rhetorical Communication, Prentice Hall, Englewood Cliffs, NJ.

McCroskey, L.L., Teven, J.J., Minielli, M.C. and Richmond Mccroskey, V.P. (2014), “James C. McCroskey’s instructional communication legacy: collaborations, mentorships, teachers, and students”, Communication Education, Vol. 63 No. 4, pp. 283-307.

McQueen, R. (2012), Communication Competence and Apprehension Within the Learning Community Context, master’s thesis, California State University, Sacramento, CA.

MacGregor, J., Tinto, V. and Lindblad, J.H. (2000), “Assessment of innovative efforts: lessons from learning community movement”, in Suskie, L. (Ed.), Assessment to Promote Deep Learning, American Association of Higher Education, Washington, DC, pp. 41-48.

Mahoney, S. and Schamber, J. (2011), “Integrative and deep learning through a learning community: a process view of self”, The Journal of General Education, Vol. 60 No. 4, pp. 234-247.

Mamerow, G.P. (2014), “Three models of learning community that contribute to participants’ learning experiences and outcomes”, Order No. 3620707, The University of Wisconsin – Madison, ProQuest, Ann Arbor, MI.

Mankiw, N.G. (2015), Principles of Economics, Cengage Learning, Stamford, CT.

Mull, L.J. (2006), Social Anxiety and Introversion in College Students, doctoral dissertation, Pacific University, Forest Grove, Oregon.

Richmond, V.P. and Mccroskey, J.C. (1997), Communication: Apprehension, Avoidance, and Effectiveness, Prentice Hall, Upper Saddle River, NJ.

Schroder, R. (2010), “Teaching across disciplines: using collaborative instruction in undergraduate education”, Journal of Economics and Finance, Vol. 34 No. 4, pp. 484-488.

Siegesmund, A. (2016), “Increasing student metacognition and learning through classroom-based learning communities and self-assessment”, Journal of Microbiology & Biology Education, Vol. 17 No. 2, p. 204.

Smith, B.L. and Macgregor, J. (2009), “Learning communities and the quest for quality”, Quality Assurance in Education, Vol. 17 No. 2, pp. 118-139.

Tinto, V. (2003), “Learning better together: the impact of learning communities on student success”, Higher Education Monograph Series, Vol. 1 No. 8, pp. 1-8.

Williams, J. (2006), “Assertion-reason multiple-choice testing as a tool for deep learning: a qualitative analysis”, Assessment and Evaluation in Higher Education, Vol. 31 No. 3, pp. 287-301.

Wingert, J.R., Karin Peterson, S.A.W., Mathews, L.G., Amy, J. and Lanou, D.C. (2014), “The impact of integrated student experiences on learning”, Journal of the Scholarship of Teaching and Learning, Vol. 14 No. 1, pp. 42-58.

Wingert, J.R., Wasileski, S.A., Peterson, K., Mathews, L.G., Lanou, A.J. and Clarke, D. (2014), “The impact of integrated student experiences on learning”, Journal of the Scholarship of Teaching and Learning, Vol. 14 No. 1, pp. 42-58.

Zhao, C.M. and Kuh, G.D. (2004), “Adding value: learning communities and student engagement”, Research in Higher Education, Vol. 45 No. 2, pp. 115-138.

Further reading

Levine, J.H. and Shapiro, N.S. (2000), “Curricular learning communities”, New Directions for Higher Education, Vol. 2000 No. 109, pp. 13-22.

Moller, L. and Soles, C. (2001), “Myers Briggs type preferences in distance learning education”, International Journal of Educational Technology, Vol. 2 No. 2.

Navidi, W. and Monk, B. (2016), Elementary Statistics, McGraw-Hill Higher Education, New York, NY.

Pashler, H., Mcdaniel, M., Rohrer, D. and Bjork, R. (2008), “Learning styles concepts and evidence”, Psychological Science in the Public Interest, Vol. 9 No. 3, pp. 105-119.

Corresponding author

Kara Smith can be contacted at: kara.smith@belmont.edu