Study Criteria

We emailed an invitation to complete a Web-based survey to everyone at Southern Illinois University School of Medicine (SIU-SM) (2,862 people). Respondents were introduced to the idea of synchronicity and were notified that the purpose of the study was to elicit their opinion on their experiences with coincidences. Respondents were informed that the questionnaire would be available for 6 days and would take 5 to 10 minutes to complete. Voluntary, anonymous participation was emphasized.

The survey included a prototype coincidence story as published by Coleman ² and the WCS-2. ³ The survey consisted of demographic information (age, gender, ethnicity, and religious affiliation) and a question regarding whether they worked in direct patient care. Respondents were also asked about their experiences with perceived coincidences around positive (eg, marriage, birth of a child, promotion, achievement, vacation) and negative life events (eg, injury, illness, death, divorce, financial stress, stress at work, school). ⁴

Participants were asked to rate WCS-2 questions such as “I think of calling someone, only to have that person unexpectedly call me” on a five-point Likert scale from “never” to “very frequently,” and questions such as “I believe that human minds are interconnected” from “strongly disagree” to strongly agree.” The response choices for all variables ranged from 1 to 5.

A total of 286 people responded, a 10% response rate. Sample size calculations found the number of respondents marginally acceptable (90% confidence level; 4.6% margin of error). Post-hoc power analyses indicated significant power (1.00) for correlational analyses (effect size 0.05; alpha error 0.05). All analyses were conducted using Predictive Analytics Software (PASW) 18. Frequencies and descriptive statistics were obtained for all variables. Principal component analysis was used to create the WCS-2 item scales. Relationships between WCS-2 scales and age, direct patient care, positive events, and negative events were assessed using Spearman’s rank correlation coefficients.

Results

Most respondents (n = 198; 72.3%) were female. Ages ranged from 18 to 74 years; the most represented age ranges were 25 to 34 years (n = 92; 33.3%) and 45 to 54 years (n = 85; 30.8%). Most were white (n = 236; 86.4%), with small percentages black, Asian, Hispanic, or other. More than half of respondents were Christian; 39.3% (n = 108) Protestant; and 19.3% (n = 53) Roman Catholic. Other faiths (eg, Buddhist, Hindu, Jewish, and Muslim), as well as agnostics and atheists, were represented in small numbers. About half (n = 131; 47.6%) work primarily in a direct patient-care position.

The means and standard deviations for the WCS-2 and positive and negative association questions are reported in Table 1 (see page 574). The lowest reported WCS-2 mean (2.11) was for the statement: “I experience strong emotions or physical sensations that were simultaneously experienced at a distance by someone I love.” The WCS-2 question with the highest mean (3.64) was “I believe that God speaks to us through meaningful coincidences.” Positive and negative association questions had reasonably high means — 3.68 and 3.33, respectively. All variables are normally distributed. ⁵

PCA factor analysis was conducted to assess whether this data also factored into three scales, as indicated by prior literature. ¹ The WCS-2 variables factored into three scales: interpersonal, agentic, and analysis/interpretation (Table 2, see page 575). One analysis/interpretation item — respondents’ beliefs that coincidences can be explained by laws of probability or chance — did not fit (factor loading -.657). It was analyzed as a separate construct, called probability/chance. The means and standard deviations for the interpersonal and agentic scales were similar to prior published results. The analysis/interpretation scores were considerably higher than prior literature has reported. ³ The reliability of our scales was high, with Chronbach’s alphas of 0.801 for the interpersonal scale; 0.827 for the agentic scale; and 0.870 for the analysis/interpretation scale.

Table 2. Means, Standard Deviations, and Chronbach’s Alphas of WCS-2 Scales

There was only one significant relationship between age or direct patient care and the WCS-2 scales. Direct patient care was negatively related to the interpersonal scale (Spearman’s rank correlation coefficient= -0.130, P = .035). However, age and probability/chance were negatively related (Spearman’s rank correlation coefficient = -0.227, P < .001), while direct patient care and probability/chance were positively related (Spearman’s rank correlation coefficient = 0.129, P = .044).

All correlations between positive or negative life events, the three WCS-2 scales, and probability/chance were significant, ranging from 0.362 to 0.471 for positive events and -0.174 to 0.303 for negative events (P < .01). Additional t-tests indicated that those who associate coincidences with positive and/or negative life events differ on all three WCS-2 scales (higher means) and the probability/chance variable (lower mean), compared with those who don’t (P = .001) (see Table 3, page 575).

Table 3. Spearman’s Rank Correlation Coefficients Between the WCS-2 Scales and Probability/Chance, and Positive and Negative Life Events

Discussion

The purpose of this study was to describe the coincidence experiences in a population associated with a medical school. We also wished to assess the possible relationship between coincidences and age, direct patient care, and positive and negative life events. The study was also useful in assessing the psychometric properties of WCS-2.

Coincidence experiences are common among those working at the SIU-SM. Overall, the means for the interpersonal and agentic scales are comparable to previously published results. ³ Our analysis/interpretation scale — which differed from the original WCS-2 scale in that the probability/chance variable did not load in the factor — was significantly higher than what Coleman and Beitman reported. ³ This may be related to nuances among the respondents, potentially due to their work in the health care field. Our three scales were highly reliable, with Chronbach’s alphas of 0.801 for the interpersonal scale, 0.827 for the agentic scale, and 0.870 for the analysis/interpretation scale. These alphas are all stronger than previously reported, ³ potentially suggesting that the WCS-2 is more reliable among those working at a medical school. Chronbach’s alpha for the analysis/interpretation scale reduces to 0.677 when the probability/chance variable is included.

It was suggested previously that age might influence coincidence detection. ³ The hypothesized explanation is that the search for meaning, which is associated with coincidence experiences, is more likely to characterize elderly people. ³ Our respondents were mostly aged 25 to 64 years, so it is possible that there were not enough older respondents to find a relationship. Only one relationship was found between age and any of the coincidence variables. Age and probability/chance were negatively related; as age increases, agreement with the statement, “I believe coincidences can be explained by the laws or probability of chance” tends to decrease. Considering this was the only finding regarding age and coincidence detection, replication studies with larger, more representative samples are needed to clarify the true relationship.

We also assumed that working in a direct patient-care position might subject respondents to higher-intensity emotional experiences compared with people working in nonclinical settings. Direct patient care was related negatively to the interpersonal scale, suggesting that direct care workers tend to have fewer coincidences involving relationships with other people or coincidental experiences involving media. Direct patient care was also related positively to the probability/chance variable. Overall, these findings may suggest that the unique experience of working in direct patient care might encourage these individuals to have a more cynical view of life and human relationships.

Prior work has indicated that people are more likely to experience coincidences at times of significant emotional involvement. ^3,6 Certain characteristics are associated with high incidence of coincidence, including vitality and negative affect, which may influence the perceived emotional meaning of an event. ³ We explored the relationship between positive and negative life events, such as marriages, births, deaths, or divorces, which are top stressful life events. ⁴ We found small to moderate, but significant, relationships between these life events and coincidence detection. These findings align with previously published results. Relationships were stronger in times of heightened emotional experiences associated with positive life events, suggestive that people find more meaning during positive, rather than negative, times in their lives.

According to Coleman and Beitman, one of the limitations of coincidence research is the WCS-2 scale itself. ³ Our analysis indicates that the three WCS-2 subscales stand as different constructs with one outlier: the belief that coincidences can be explained by laws of probability or chance. For that reason, it was removed from the analysis/interpretation scale and analyzed separately. People who experience coincidences tend not to believe they are the result of chance, which aligns with the “search for meaning” hypothesis. ³

This project has a number of limitations. Although the number of respondents was acceptable for analysis, the total response rate was low. The responses may overrepresent those who were interested in the topic of coincidences. The respondents also tended to be female and white, limiting the generalizability of our results. Further, it would have been helpful to know whether those reporting direct patient care were doctors, nurses, other allied health professionals, or clinic receptionists.

Conclusion

Our findings indicate that SIU-SM affiliates commonly perceive coincidences. The WCS-2 items factored into three scales, with one outlier. To improve inter-item reliability for future researchers, we suggest that the probability/chance variable be analyzed separately. There is a limited relationship between coincidence detection and both age and direct patient care. However, the WCS-2 scales were related to both positive and negative life events.

Coincidence detection around positive and/or negative life events suggests an association with emotional states and a search for meaning that aligns with previously published findings. Future researchers may wish to assess the relationship between religion and coincidence detection. Previously, an overlap between spirituality and coincidences was described, with those who report more coincidences also tending to report more spiritual experiences. ²

We also suggest exploring the influence of high-intensity clinical settings, such as the ED, versus milder health care settings, such as outpatient clinics. Researchers should strive to collect data that would allow multivariate models, which could control for a number of variables pertinent to daily life and perceived coincidences.