The “Ghost” in the Lab: Believers’ and Non-Believers’ Implicit Responses to an Alleged Apparition

ABSTRACT

Many nonbelievers may engage in supernatural thinking despite their statements to the contrary. Using belief in the afterlife as a test case, we examine, across two studies, the possible discrepancy between what people say they believe and how they reason implicitly. In Study 1, participants completed a mindfulness task during which a light went off unexpectedly. Half had previously been told that a ghost had recently been seen in the same room. Participants’ electrodermal responses and heart rate variability suggested implicit attributions to the “ghost,” and these physiological effects were unrelated to afterlife beliefs. In Study 2, compared to those in a control condition, participants who were informed that a ghost had been seen in the laboratory chose to sit further away from the alleged apparition. Surprisingly, this distancing was most pronounced among participants who did not believe in the afterlife. Cumulatively, the data indicate that self-report measures of supernatural belief may not fully capture private experience and implicit reasoning.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ Heart rate variablity measures were initially selected based on Delaney and Brodie (2000), excluding those inappropriate for ultra short-term measurement (according to Shaffer & Ginsberg, 2017: SDNN, PNN50, VLF Power, and LF Power). These included the root mean square of successive differences (RMSSD), low frequency/high frequency ratio (LF/HF ratio), normalized low frequency power (LF power nu), normalized high frequency power (HF power nu), and Heart Rate. However, it later came to our attention that LF power nu, HF power nu as well as and LF/HF ratio have questionable validity (Billman, 2011; Goldstein et al., 2011) as indices of sympathetic, parasympathetic activity and sympathetic/parasympathetic balance, respectively. Thus we decided to report only heart rate and RMSSD.

² When the five non-naïve participants were excluded, the only significant effects were of time, χ² (1) = 31.23 p < 0.01 and condition X time interaction χ² (1) = 5.89, p = 0.01.

³ When the five non-naïve participants were excluded the main effect of time is significant χ²(1) = 5.40, p = 0.02, as well as the condition x time interaction χ²(1) = 5.58, p = 0.02. All the other effects were not significant.

⁴ When the five non-naïve participants were excluded, the only significant effects were of condition χ²(1) = 6.51, p = 0.01 and time, χ²(1) = 7.77, p < 0.01.

⁵ When the non-naïve participants were excluded, the odds for choosing to sit closer to the door for the “ghost” group were 0.38 times that of participants from the control group 95%CI[17, 85].