Abstract
Background
Creativity involves the generation of novel ideas that are original and unique. It is a subjective process, and few studies are available in support of objective measures. Available tests of creativity are limited to questions related to an individual’s trait and subjective responses. Though creativity is a divergent construct, an objective approach to computing and marking one as creative is required. This is so because creativity is an important factor for success, and a subjective approach would bring bias.
Purpose
The present study aims to develop a creativity score using the Rorschach Inkblot Test (RIBT) and then test it with eye-tracking technology.
Methods
Thirty-four university students were recruited for the study using a purposive sampling technique. RIBT cards were shown on a computer screen with an eye tracker mounted on it. Their responses were recorded and analysed to develop a novel construct of the Creativity measure. The Creativity score is then divided into high, medium and low creativity using the k-means clustering algorithm. Eye parameters of fixations, saccades and pupil diameter were explored for each group.
Results
ANOVA revealed significant differences between the three groups. In the high-creativity group, fixation count, variations in pupil diameter and total saccadic duration were higher than their counterparts. Mean fixation duration was highest for the low-creativity group.
Conclusion
The results indicated that using unstructured blots with Eye-tracking technology helps assess creativity objectively, further broadening avenues to measure creativity.
Keywords
Introduction
Creativity can be defined as the production of novel, valuable and relevant ideas. 1 It has been defined in various ways, such as a type of intelligence, 2 an unconscious process, 3 part of problem-solving, 4 and as a means of forming new associations. 5 Many theorists have also emphasised that the production of creativity involves producing original and rare ideas.6, 7
Measurement of Creativity is a topic of concern, as the definition varies, so the techniques. In the current scenario, mostly used technique is derivation of Creativity Quotient, which can be obtained from tests such as Torrance Tests of Creative Thinking and Guilford’s Alternative Uses Test. Other measures are self-rating scales (Kaufman Domains of Creativity Scale and Biographical Inventory of Creative Behaviours) and social personality tests such as ‘Openness to Experience’ in NEO Big Five Personality Traits. 8 Projective techniques, such as the Rorschach Inkblot Test (RIBT), stimulate open and unique thinking, making them effective for assessing creativity.9, 10
Although these methods are widely used, the question of reliability comes due to subjective judgements and bias in reporting.11, 12 In these circumstances, a valid physiological measure of creativity would be a more reliable approach in assessing it. If conducted in conjunction with a psychological test, a more comprehensive measure will be obtained. In this study, we will be using RIBT in conjunction with eye movements (EMs) to identify creativity.
Application of RIBT in Creativity
Hermann Rorschach explained the validity of RIBT in differentiating between individuals with different levels of creativity. It is highly correlated with the Torrance test of creativity. 13 Certain RIBT metrics are highly associated with creativity, such as Movement (M), Originality (O), Whole locations (W), Popular (P), and elaborative responses (known as ‘Descriptors’ in this paper). 14–16. Other Rorschach markers include good form quality (F+), total responses (R), and color-form combinations. 16
Eye Movements in Creativity
Eye movement (EM) analysis helps to understand the creativity in real-time by capturing visual behaviour and its associated cognitive processes, such as attention. 17 Studies identified creativity with distinct EM patterns. To explain this, the associated cognition is to be identified, that is, during creative ideation, attention shifts from External Directed Cognition (EDC) to Internal Directed Cognition (IDC). IDC is manifested as having longer fixations, high variability in pupil size and at times aversion from the stimulus.18, 19 Eye aversion is frequently linked to visual imagination and problem solving.20, 21 In his study, Ueda explained that creative individuals exhibit high saccadic frequency, more eye blinks and reduced micro-saccades during creative tasks. 19
When this information on EMs can be blended with psychometric tools, a more insightful approach is taken in studying creativity.22, 23 In this study, we hypothesise that creativity can be measured and categorised using a valid psychometric tool (here, RIBT). Also, EMs or Eye Parameters (EPs) will be different in each group.
The study aims to categorise creativity levels using RIBT metrics from the literature. Then, the EMs of each group will be measured to find if there are any significant markers associated with each group.
Methods
Sample
The sample consisted of 34 participants (9 females and 25 males) with a mean age of 26.76 years (SD = 3.59). These participants were post-graduate students from a residential university campus, selected using purposive sampling. The inclusion criteria consist of: (a) none of the participants had prior exposure to the RIBT, (b) all reported corrected or near-corrected vision, (c) no history of psychiatric disorders or substance abuse in the last 6 months and (d) no history of eye-related disease or disorder. Prior to the study, all participants were assessed for anxiety and general pathology using the State-Trait Anxiety Inventory (STAI) and the General Health Questionnaire (GHQ).
Procedure
A description about the study was briefed to the participants. If they agreed, consent was taken from the students to participate in the study, following the Helsinki declaration. To screen out anxiety issues and check general pathological conditions, STAI and GHQ were administered. Once the experiment was explained to the participants, their eyes were calibrated via an eye-tracking device (Tobii X3-120 device with 120 Hz sampling frequency), which was mounted at the bottom edge of a 21-inch monitor. Then, RIBT cards were shown on the computer screen and the whole experiment was conducted in the presence of a licensed clinical psychologist. The clinician recorded the participants’ verbatim as a written note. The eye-movement data and video recording were acquired along the same time. Bruno Klopfer protocol for RIBT was followed for data collection and quantitative scoring of the responses. 19 The responses were coded by three participants for inter-rater reliability. For analysing EPs, Tobii Software was used for retrieving raw data of gaze plots and time stamps, and then, it was run using developed code to get fixations, saccades and pupillary changes. The detailed experimental design is provided in Figure 1.
Design of the Experiment.
Tools Used
State-Trait Anxiety Inventory
STAI is developed assess state and trait anxiety levels. It has two forms: Y-1 state anxiety, which purports to measure how one feels anxious at the present moment, and Y-2 trait anxiety, which evaluates one’s general anxiety level. The scale has internal consistency coefficients ranging from 0.86 to 0.95. The test–retest reliability coefficients have ranged from 0.65 to 0.75 over a 2-month interval with high construct and concurrent validity. 24
General Health Questionnaire
GHQ is a 28-item questionnaire that measures somatic symptoms, anxiety, insomnia, social dysfunction and severe depression. 25 Pathology is measured when individuals score above the cutoff score. The scale has a test-retest reliability of 0.78–0.9 with excellent inter-rater and intra-rater reliability. 26 It correlates well with the Hospital Depression and Anxiety scale. 27
Eye Tracker
An eye tracker is a gadget that tracks EMs and gaze points using illumination, sensors and processing. Regardless of the ambient light, precise and continuous tracking is made possible using near-infrared light. Pupil centre corneal reflection eye tracking is commonly used to describe this technology. It collects information about eye position, gaze direction and EMs with extreme precision using optical sensors. Most are predicated on the core idea of corneal reflection tracking. 28 Here, in this study, the Tobii X-3 120 Hz static eye tracker is used; this has 0.5-degree accuracy and has been used successfully in research studies.23, 29
Rorschach Inkblot Test
Hermann Rorschach empirically developed the test in 1921. This tool is used to assess the personality disposition of the subjects in terms of preferential response style, stress tolerance capacity and accessibility of coping resources. Projective techniques are based on the projective hypothesis—people may respond to a vague or ambiguous situation is often a projection their underlying feelings and motives. It comprises five chromatic cards (2, 3, 8, 9, 10) and five achromatic cards (1, 4, 5, 6, 7). The cards are numbered on the back serially.
Among the various established interpretive techniques of RIBT, the Klopfer system would be helpful in deducing the creativity of an individual due to its emphasis on qualitative scoring in addition to quantitative. 30 The system emphasises imaginative responses, which are highly associated with creativity. 31 In addition, compared with other methods, Klopfer allows for greater interpretation flexibility, like using additional responses, and is comparatively flexible in approach.32–34 The next section highlights the derivation process from both tools, as explained in Figure 2.
Data Analysis.
Derivation of Creativity score
RIBT Metrics
Five highly occurring quantitative metrics derived from the literature were used to measure Creativity. These metrics were Movement (M), Descriptors (D), Whole location (W), Popular (P) and Original (O) responses. These were computed for each participant from their responses. Movement was marked where there was animation in the responses elicited from the images. These include human, animal and inanimate movements. Descriptors were the features given by the participants to explain the response. Whole location means that the responses they have provided utilize the whole blot of the card. Popular responses were those that were mentioned in the Klopfer protocol as Popular or most frequently occurring responses. Originality was marked only when a participant from the sample had given a response that was unique in terms of its content and location. When a response was given by more than one participant and was not a popular one, then they were marked as cluster responses (Figure 3).
Development of Response Type.
Evaluation of the Metrics to Development of the Creativity Score
Let there be 10 numbers of cards.
The following metrics were acquired for participant 1,
Movement (M) responses (each cards): {A1, A2, …, A10}
Descriptors (D) responses (each cards): {B1, B2, …, B10}
Whole (W) responses (each cards): {C1, C2, …, C10}
Popular (P) responses (each cards): {D1, D2, …, D10}
Original (O) responses (each cards): {E1, E2, …, E10}
Number of responses (each cards): {R1, R2, …, R10}
Similarly, we computed ratios for D, W, P and O metrics.
Then, each of the responses was normalised using the following equation:
where X = Ratio of each metric,
Maximum = Maximum value in the sample,
Minimum = Minimum value in the sample.
This was done to remove outliers in scores and normalise them in the sample from the data obtained. Since there were five metrics, an equal weightage of 0.2 was given to keep the data within the maximum value of 1. Then, the total of all the metrics was computed for each of the participants. The summation is the Creativity score.
Categorisation of Creativity Score
The scores were then categorised as per k-means cluster sampling. The data were then divided into three groups of high, medium and low levels of creativity (Figure 4).
Derivation of Creativity Scores.
Derivation of Eye Parameters
The second step was to develop an algorithm for analysing EPs obtained during the study. EPs related to fixation, saccades and pupil diameter profiles were derived from Tobii Pro Software and then filtering of the data. The following EPs were derived for each creativity level, as represented in Figure 6.
Statistical Analysis
IBM SPSS Statistics Software 23 was used to run the statistical analysis for exploring the significant eye parameters to distinguish the three levels. These findings would also help in obtaining the eye movement patterns in each group.
Results
Creativity score was derived and categorised into three levels of namely high, medium and low (Figure 6). As per the clustering method, the demarcations of the obtained score came as follows: 0–0.34 for the low, 0.35–0.48 for the medium and 0.49–0.7 for the high-creativity group.
EPs were computed for each participant on all Rorschach cards (Figure 5). One-way ANOVA was applied to these parameters, keeping the derived creativity score as a factor and EPs as the dependent variable. The significance level was decided at a 95% confidence interval. The statistical variations found for different EPs are explained in Table 1.
Extracted and Deduced Relevant Eye Parameters to Measure Creativity.
Creativity Level as Per k-means Sampling.
Mean Differences Between the Creativity Levels in Each Type of Eye Parameters.
The corresponding graphs illustrate the descriptive statistics in Figure 7A–K. Here, the vertical axis represents the ‘mean of eye parametric type’, and the error bar indicates the standard deviation for each creativity level.
Discussion
The main purpose of the study is to categorize participants into different levels of creativity using Rorschach metrics and then to find the significant eye pattern that would distinguish the three groups. To serve this purpose, a creativity score was developed using scores from five metrics and then divided into three levels using k-means sampling. It is represented in Figures 4 and 6. Then, for each group, eye parameters were examined statistically. Results indicate significant differences among the three creative groups in their EM patterns (Table 1). We will discuss each of the eye parameters in separate sections.
Fixation Profile
Participants in the high creative group showed a greater number of fixations than their counterparts, as can be seen from Figure 7A. This indicates they demonstrated increased engagement while visualizing these cards. 35 Meanwhile, individuals in the low creative group fixated on the cards for a longer period as indicated by their higher Mean Fixation Duration (MFD), shown in Figure 7B. High MFD indicates increased use of cognitive resources. 35 This suggests that highly creative individuals possess a greater ability to recognize and interpret complex visual stimuli in less time.36, 37 Also, from the Rorschach metrics, it is seen that the high-creativity group gave more responses with descriptions than is seen in other groups.
Pupil Diameter
The medium-creativity group reflected the largest Mean Pupil Diameter for both eyes (Figure 7F). Skewness of Pupil Diameter showed symmetrical distribution across all creativity levels, with the high-creativity group demonstrating the largest Mean Absolute Change in Percentage of Pupil Diameter, reflecting greater variability in pupil size during visual processing (Figure 7H), consistent with findings by Benedek. 18 The high-creativity group also showed a significant pupil dilation and constriction, which indicates more dynamic processing of visual stimuli and a feature of cognitive flexibility. 38
Saccades
The medium-creativity group had the highest Skewness of Saccadic Duration, represented in Figure 7J, while the high-creativity group displayed the longest Total Saccadic Duration (TSD), shown in Figure 7K. Highly creative individuals frequently shift focus between different elements of stimuli, as reflected in their saccadic patterns. 39
Graphical Representation of the Changes in Eye Parameters, Illustrated Through Descriptive Statistics.
Conclusion
This study aimed to explore creativity through established RIBT metrics. We found some intriguing patterns in eye-tracking measures across the three creativity groups. In the high-creativity group, there were higher FCs, greater Mean Absolute Change in Pupil Diameter Percentage and longer TSD, with lower Skewness in Pupillary Changes. Meanwhile, the low-creativity group showed longer MFDs when viewing the abstract RIBT images. These findings suggest that the RIBT and eye-tracking metrics can offer meaningful insights into creativity and open a new assessment avenue.
Also, the study has some limitations. This is because we relied on existing literature and selected five core metrics to derive creativity scores, but selection bias is possible. Even so, the study met its main objective of exploring creativity through these responses. The observed differences in EPs lend support to this approach. In future studies, using a larger sample size and an additional validated creativity scale could help to find subtle distinctions between the creativity groups.
Abbreviations
AUT: Guilford’s alternative uses test; BICB: Biographical inventory of creative behaviours; CQ: Creativity quotient; D: Descriptors; EDC: External directed cognition; EM: Eye movements; EP: Eye parameter; F+: Form quality good; K-DOCS: Kaufman domains of creativity scale; M: Movement; O: Originality/original responses; P: Popular responses; R: Responses; RIBT: Rorschach inkblot test; TTCT: Torrance tests of creative thinking; W: Whole locations.
Authors’ Contribution
The first author, SN, collected and analysed data on RIBT metrics and statistics and edited the article. The second author, AKR, used a k-means clustering algorithm to divide the groups, analysed the eye parameters and contributed to making flowcharts and graphs related to the article. RG, the third author, provided guidance, support and an environment for experimenting. All the authors have given their consent for the article.
Statement of Ethics
The study has received ethical approval (No. IIT/SRIC/DR/2019) from the Indian Institute of Technology’s ethical committee (Kharagpur, India).
Footnotes
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
Funding
The authors received no financial support for the research, authorship and/or publication of this article.
Patient Consent
Participants were given verbal information regarding the study, and written consent was obtained. The consent includes the approval to record their audio and video and to use it in research and publication.
ICMJE Statement
All four criteria of ICMJE for authorship have been met for this study.
