Critical thinking

Meaningful learning is necessary for critical thinking. The operational definition of critical thinking is a metacognitive, nonlinear process of purposeful judgment that includes self-directed learning and self-assessment[13, 14]. How critical thinking should be taught and how it is learned are unclear,[15, 16] especially at the medical school level. Willingham[15] stated that critical thinking occurs when a student penetrates beyond the surface structure of a problem and recognizes how the problem can be solved, and in addition, possesses the content knowledge integral to solving the problem. Without both components, a student may be able to critically analyze one problem, but will falter when given a similar problem in a different context[15]. Graduating physicians should be able to critically evaluate novel cases that they encounter in the clinic using their previous, albeit limited, clinical experiences[17].

Concept mapping in medical education

In graduate medical education, West et al[17] used the concept map learning strategy developed by Joseph Novak[18] in resident physicians, and studied the validity and reliability of concept mapping assessment (CMA). They found that concept maps could be scored reliably and CMA could measure changes in the conceptual framework of physicians[17].

Mind mapping in medical education

Mind mapping was developed by Tony Buzan[19] and the inspiration for this strategy arose from the notebooks of Leonardo da Vinci[20]. Mind maps, like da Vinci's notes, are multi-sensory tools that use visuospatial orientation to integrate information, and consequently, help students organize and retain information[21, 22].

Mind maps can be used as a teaching tool to promote critical thinking in medical education by encouraging students (adult learners) to integrate information between disciplines and understand relationships between the basic and clinical sciences[21]. The ability to integrate information by finding valid relationships between concepts allows students who construct either mind maps or concept maps to reach a metacognitive level[15]. However, the added dimensions of pictures and colors that are unique to mind maps have not only been shown to facilitate memory,[23] but may appeal to a wide range of students withvisual- and linear-oriented learning styles. Consequently, the advantage of using mind maps in medical education is that this strategy may benefit more students with diverse learning styles.

Both mind maps and concept maps allow students to recognize the intra- and inter-relationships between concepts, which reflects the kind of real-world thinking predominant in the clinical setting[24].

Farrand et al[25] were the first group to investigate the potential role of mind mapping in medical education. These researchers explored whether the mind map learning technique was superior to traditional note taking in both short- and long-term factual recall of written information in medical students. They found that the mind map technique significantly improved long-term memory of factual information. Additionally, they found significant differences in self-reported motivation with the mind map group having lower levels of motivation than the self-selected study group. Although not supported by other literature, this finding may be explained by the fact that students were not given adequate time to adjust to using the mind map technique, and therefore, may have felt less comfortable using it. Although the results of the study were promising, the authors did not address critical thinking. Consequently, studies exploring the relationship between mind mapping and critical thinking are needed before the usefulness of mind mapping can be fully supported in medical education.

Wickramasinghe et al[26] were the second group to investigatethe effectiveness of mind maps in medical education. Using a similar study design as that used by Farrand et al,[25] these authors assigned new entry medical students into 2 groups: mind map and self-selected study groups. The authors also developed a method to score the mind maps based on structure and content; however, they did not describe the method nor did they provide any data to support it[26]. The authors reported that there was no significant difference in scores between groups[26]. They did, however, report that all of the subjects in the mind map group perceived that mind maps are useful for memorizing information. Based on their findings, the authors concluded that mind mapping may not be effective in improving retention of short-term information[26].

Mind maps and concept maps

Although concept maps and mind maps have similar characteristics, they are fundamentally different in design. Concept maps are devoid of color and pictures, and are constructed in a top-to-bottom hierarchy. Mind maps, in contrast, use a central theme in the middle of a page with categories and subcategories that radiate peripherally, thus making them truly non-linear. The cross-links among categories highlight their intrinsic relationships, and allow the student to compare and contrast information. Unlike concept maps, mind maps are multisensory--they include color and pictures, which facilitate the conversion of information from short- to long-term memory[23, 27]. An example of a mind map created by a medical student in this study can be found in Figure 2.

Since critical thinking is dependent upon both content (domain) knowledge and problem familiarity,[15] mind mapping may facilitate critical thinking because it fosters student retention of factual information, as well as relationships between concepts[25]. Currently, however, there are no data to support the hypothesis that mind maps facilitate critical thinking in medical students.

Purpose of the study

The primary purpose of this study was to investigate whether a relationship existed between the mind map learning strategy and critical thinking, as measured with the Health Sciences Reasoning Test (HSRT), and whether this relationship was stronger than one between the preferred learning strategy of standard note-taking (SNT) and critical thinking.

The secondary purpose of this study was to determine whether mind maps were superior to SNT in the short-term recall of factual information. Mind map depth was assessed using the previously published Mind Map Assessment Rubric[28].

Study setting and sample

After full approval by an Institutional Review Board, this study was conducted during the 2008-2009 academic year at a US medical school located in a large metropolitan area.

An a priori power analysis[29] using a one-tailed t-test revealed a minimum sample size of 70 subjects. This calculation was based on the following: effect size d = 0.8, alpha = 0.05, and power = 0.95. The large sample size (N = 131) assumes a normal distribution of the population, and therefore, parametric statistics were used to analyze the data. The sample of convenience consisted of first-year medical students who voluntarily participated in this study.

Procedures

The independent variable in this study was the note-taking strategy used by the medical students. Subjects were randomly assigned to 2 note-taking groups: a standard note-taking (control) group and mind map (experimental) group. The design of the study is outlined in Figure 3.

Subjects in both note-taking groups were asked to learn information contained in a 394-word text passage—on the topic of cacti and other succulent plants—from the verbal ability section of a previously published Graduate Record Examination (GRE). This topic was chosen to reduce the chance that the medical students would have previous advanced knowledge of this field. The GRE is a standardized entrance examination used as part of the US graduate-school admissions process. The exam is used by faculty to decide which students will be admitted to graduate school and who will be awarded academic fellowships. A GRE text passage was used in this study because the GRE is taken by students who are, in general, of a similar age to those entering US medical schools. Consequently, the text passage was at an appropriate cognitive level for medical students. A post hoc analysis of the medical students in the study revealed that none of them majored in botany at the undergraduate level.

Subjects in the control group used standard note-taking (SNT) strategies that they used throughout their academic careers to learn the text passage. SNT is defined as any study strategy that does not rely on reorganizing information using architecture commonly seen in a concept map or mind map[25]. SNT is a process whereby notes are arranged in a hierarchy from the top of a page to the bottom, or from left to right, without any hierarchy[30]. Subjects in the experimental (mind map) group were given a 30-minute presentation on mind maps and then instructed to create mind maps in order to take notes on the material in the text passage.

There were two dependent variables in this study. The first one was the score on the text passage quiz, of which there were two. These two quizzes, which were based on the content of the GRE text passage, were administered to all subjects after assignment to the groups. All subjects were simultaneously (but in different rooms) exposed to the passage for 5 minutes and were not permitted to write any notes. The passage was collected and followed by the administration of math quiz 1. This quiz was used to "blank" the minds of the subjects by preventing the simple recall of information that could result in a higher quiz score and confound the results[25].

After math quiz 1, all subjects were administered text passage quiz 1. The purpose of this 5 multiple-choice question quiz was to test the students' factual understanding of the passage without any note-taking strategy. This baseline quiz was used as a covariate to account for potential differences between the groups prior to initiating any note-taking strategy.

After taking text passage quiz 1, subjects in the mind map group were given a presentation on mind maps and how to construct them, while at the same time, subjects in the control group were sequestered for a break and could not leave the lecture hall. After 30 minutes, all subjects were then re-exposed to the text passage and instructed to take notes using either standard note-taking (SNT) or mind maps (MMs), depending on their group assignment. All subjects were given 25 minutes for note-taking and at the end of this time period, all passages and notes were collected. This was followed by the administration of math quiz 2 in order to again discourage the simple recall of information by the subjects. After math quiz 2, all subjects were simultaneously administered text passage quiz 2 based upon the passage. This quiz consisted of 10 multiple-choice questions: the same 5 questions from quiz 1 plus an additional 5 questions. This was done to see if the students retained the factual information and to address potential testing effects (ie, higher scores due to repeated testing exposure).

The second dependent variable of this study was the HSRT score. The HSRT consists of 33 multiple-choice questions that measure critical thinking by challenging students to form reasoned judgments based on textually presented information consisting of a number of vignettes[31]. The information presented in the vignettes includes diagrams, charts, and other data related to health care scenarios. The HSRT does not test domain knowledge (ie, subject-specific knowledge such as that found in anatomy and biochemistry); therefore, subject-specific knowledge is not needed by the students taking the exam. The HSRT has been extensively studied in health professional students and working professionals[14, 31].

The HSRT reports an overall numerical score and 5 subscales: analysis, inference, evaluation, deductive reasoning, and inductive reasoning. The operational definitions of these subscales, adapted from a previous Delphi study, [14] follow: analysis (ability to identify the intended and actual inferential relationships among statements, questions, concepts, descriptions or other forms of representation intended to express beliefs, judgments, experiences, reasons, information or opinions); inference (ability to identify and secure elements needed to draw reasonable conclusions; to form conjectures and hypotheses, to consider relevant information and to educe the consequences flowing from data, statements, principles, evidence, judgments, beliefs, opinions, concepts, descriptions, questions, or other forms of representation); evaluation (ability to state the results of one's reasoning; to justify that reasoning in terms of the evidential, conceptual, methodological, criteriological and contextual considerations upon which one's results were based; and to present one's reasoning in the form of cogent arguments); deductive reasoning (assumed truth of the premises purportedly necessitates the truth of conclusion and this includes traditional syllogisms, as well as, algebraic, geometric, and set-theoretical proofs in mathematics); and inductive reasoning (an argument's conclusion is purportedly warranted, but not necessitated, by the assumed truth of its premises and this includes scientific confirmation and experimental disconfirmation)[31].

Mind maps were scored using the Mind Map Assessment Rubric (MMAR). The interrater reliability of the MMAR is strong and has been reported to be 0.86[28]. Face validity of the MMAR has been investigated, and the entire rubric is available online (see reference [28]).

Sample characteristics

Sex and ethnicity distributions were similar in both groups as demonstrated in Table 1. The mean age of subjects in both groups was also similar. In the SNT group, the mean age of subjects was 24.45 years (SD = 3.26) and in the MM group, the mean age of subjects was 24.74 years (SD = 3.91). Using one-way analysis of variance (ANOVA), no significant difference in mean age between groups was found. Subjects in the SNT group had a mean total SAT score of 1285.71 (SD = 112.06) and those in the MM group had a mean total SAT score of 1254.46 (SD = 110.20). No significant difference in total SAT score between groups was found. In addition, no significant differences in SAT verbal and math subscores between groups were found. The mean total MCAT score of subjects in the SNT group was 27.26 (SD = 3.04) and the mean total MCAT score of subjects in the MM group was 27.05 (SD = 3.17). No significant difference in total MCAT score between groups was found. In addition, no significant differences in MCAT biology, physics, and verbal subscores between groups were found.

Quiz assessment of domain knowledge

The mean score of the pre-quiz (quiz 1) among subjects in the SNT group was 3.15 (SD = 1.22) and the mean score of the pre-quiz (quiz 1) among subjects in the MM group was 3.42 (SD = .84). A two-tailed independent samples t test revealed no significant difference between the means: t (129 df) = -1.47, p = .14.

The mean score of the post-quiz (quiz 2) among subjects in the SNT group was 7.85 (SD = 1.40) and the mean score of the post-quiz (quiz 2) among subjects in the MM group was 7.64 (SD = 1.22). A two-tailed independent samples t test revealed no significant difference in means between the groups: t (129 df) = .912, p = .36. Figure 4 is a bar chart depicting these data.

A comparison of the means of the pre-quiz (quiz 1) scores and post-quiz (quiz 2) scores between groups revealed no significant differences (SNT pre-quiz mean = 3.15, MM pre-quiz mean = 3.42, SNT post-quiz mean = 7.85, and MM post-quiz mean = 7.64). However, the difference between means of the pre-quiz (quiz 1) and post-quiz (quiz 2) scores in each group differed. In the SNT group, this difference was 4.70 (7.85 - 3.15 = 4.70) and in the MM group, this difference was 4.22 (7.64 - 3.42 = 4.22).

In order to further analyze these results and control for the fact that the quiz scores themselves were slightly skewed (ie, a long tail created by a few students who did very poorly), a standardized z score was used. A difference z score was created between the standardized quiz scores so that the degree to which the variability in each quiz affected the outcome would be the same. Unlike the quiz scores, the difference z score conforms to a Gaussian distribution as demonstrated in Figure 5. The difference z score is standardized with a mean of 0 and a SD of 1.08. On the average, subjects in the MM group had lower scores on the second quiz (-.2061 SD), while those in the SNT group increased by about the same amount (.2093 SD). This represents about two-tenths of a SD. The fact that the scores of the groups vacillated by almost the same amount is not by chance. A two-tailed independent samples t test revealed a significant difference between the means of the z score difference: t (129 df) = 2.241, p = .027.

HSRT assessment of critical thinking

ANOVA was used to compare the means of pre- and post-HSRT total scores and subscores between the SNT group and MM group. No significant differences were found among any of the pre- and post-HSRT total scores and subscores. The bar chart in Figure 6, which displays pre- and post-HSRT total scores, demonstrates no significant differences between pre- and post-HSRT total scores between groups.

HSRT assessment of critical thinking

The mean total score on the pre-HSRT for subjects in the SNT group was 23.41 (SD = 3.69) and the mean total score on the pre-HSRT for subjects in the MM group was 24.07 (SD = 3.04). This difference was not significant and this finding demonstrates that both groups had similar baseline critical thinking abilities as measured by the HSRT.

The mean total score on the post-HSRT for subjects in the SNT group was 23.47 (SD = 3.82) and the mean total score on the post-HSRT for subjects in the MM group was 23.97 (SD = 3.75). Subjects in the MM group did not score significantly different than those in the SNT group on the post-HSRT, a finding that suggests the power of mind mapping even when it was introduced to a novice group of subjects during a brief introductory session. The fact that subjects in the MM group scored worse on the post-HSRT compared to their pre-HSRT total scores could be explained by their unfamiliarity in creating mind maps or fatigue from the testing process. Additionally, requiring MM subjects to learn mind mapping may have created contextual interference that hampered short-term retention as demonstrated by the results of the post-HSRT; however, this may actually promote long-term retention as noted in the contextual interference literature[32]. Subjects in the MM group may have been so preoccupied with creating mind maps that they failed to think critically about the information. Therefore, repeated exposure to mind mapping over time may be a necessary requisite in order to better test whether the use of mind mapping increases critical thinking as measured by the HSRT.

Limitations and future research

The SNT group remained in the lecture hall during the break while the MM group was concomitantly exposed to a 30-minute mind map presentation. A potential limitation, therefore, is that during the break subjects in the SNT group could have mentally reviewed the text passage. These subjects were observed during this time and were not permitted to view the text passage. The possibility that they were able to accurately recall the text passage during the break (while the MM group listened to the presentation) is unlikely because they were exposed to the text passage 20 minutes before the break and had also taken an intervening math quiz (see Figure 3).

Because critical thinking takes time to develop, short-term changes in critical thinking was another limitation of the current study. Multiple mind-map sessions may be necessary for students to gain proficiency in the strategy before significant changes in the acquisition of domain-based knowledge and critical thinking emerge. Recently, Srinivasan et al[24] reported that concept map scores significantly increased in physicians who created concept maps on two separate occasions. They recommended that future concept map studies should allow subjects to create concept maps on multiple occasions. This may also be true of mind maps because, although not investigated in medical students, researchers have demonstrated that mind map depth increases as students gain proficiency in their construction over time[13, 30].

Future studies should be designed to allow subjects to create multiple mind maps so that they can gain proficiency in the technique. This would enable them to move from novice to expert regarding the creation of mind maps, and therefore, could ultimately allow them to emphasize critical thinking. Additionally, these studies could also measure longitudinal changes in HSRT scores as students become more proficient at mind mapping.