Discovery Lab Answers

The comparison, and how the library frames it

The most direct source retrieved for this question is Baptista and Derakhshani (2014), a point-by-point reply in the Journal of Parapsychology to a skeptical critique. One of their explicit analyses compares “reproducibility of psi experiments to reproducibility of experiments across related mainstream fields,” and their reported conclusion is that the two are “similar” ^[2]. That is the central claim to sit with — and it is a claim advanced by parapsychologists in reply to a critic, not a neutral third-party audit, so it should be read as one side of a documented exchange.

A coverage caveat up front: the passages retrieved for this question are a small, non-random slice — one meta-analytic reply, a couple of PA-membership surveys, a scoping review of a different topic, and two European Journal of Parapsychology issues. The library’s share of the published literature comparing parapsychology’s replication record to mainstream psychology’s has not been measured here, and I cannot characterize “the field” or “the literature” as a whole from these excerpts. What follows is what these specific sources say.

What the retrieved sources actually contain

On statistical power and the significant-study rate

Baptista and Derakhshani (2014) work through the ganzfeld as their worked example, drawing on the 105 four-choice studies in Storm, Tressoldi, and Di Risio (2010):

Their argument is that the modest proportion of significant studies (28.5%) is not evidence of an unreliable effect but is exactly what ~30% average power predicts — i.e., the significant-study rate is “completely consistent with past findings” ^[2]. This is a power/reproducibility argument that mainstream psychology has itself confronted in its replication debates.

On the decline effect and experimenter expectancy

Baptista and Derakhshani (2014) argue that decline effects and experimenter-expectancy effects are “far from unique to parapsychology,” citing Jonathan Schooler (2011), a professor of psychological and brain sciences, who catalogued decline-effect examples across mainstream domains in a debate over psi at Harvard ^[2]. The point is comparative: phenomena skeptics treat as red flags for psi also appear in mainstream fields.

On the file drawer and publication practices

Two threads in the sources bear on this:

• Baptista and Derakhshani (2014) note that awareness of the file-drawer problem came early to psi research, and point to preregistration infrastructure (the Koestler Parapsychology Unit Registry) as a means to limit residual publication bias ^[2].
• A PA-membership survey in the European Journal of Parapsychology (vol. 11) ^[1] found a 73% acceptance rate for parapsychology papers in non-parapsychology journals and a 44% response rate among actively publishing members, while members estimated the acceptance rate for their own papers at only ~18% — a gap between perceived and apparent acceptance. These are survey figures about publishing dynamics, not replication rates as such.

On null results within parapsychology

Irwin (2014), surveying 114 PA members, reports that 40% acknowledged having conducted a psi experiment that failed to establish a significant effect ^[3]. Irwin frames this as a challenge to the skeptical claim that parapsychologists never report failures. This is a data point about how the field handles its own nulls, adjacent to the replication question.

What these sources do not settle

• None of the retrieved passages provides a mainstream-psychology replication statistic (e.g., a large-scale reproducibility-project success rate) that you could set numerically beside a parapsychology figure. The “similar reproducibility” claim in ^[2] is asserted with reference to related fields, but the retrieved excerpts do not lay out the mainstream comparison numbers themselves.
• The strongest comparative claim here comes from a paper written specifically to rebut a critic ^[2]; the retrieved sources do not include the critic’s original article or an independent adjudication of it.

Skeptical critiques

What critics argue. Baptista and Derakhshani (2014) are responding to a Skeptical Inquirer article whose thesis they summarize as “Heads I Win, Tails You Lose; How Parapsychologists Nullify Null Results” — the argument that parapsychologists retrospectively explain away failed studies (retrospective nullification) and that selection/file-drawer bias inflates the apparent psi effect ^[2]. They characterize the selection-bias criticism as “a priori an extremely powerful one” ^[2] — i.e., they grant it is a serious objection before contesting it.

What the experimental data show. Baptista and Derakhshani (2014) respond that awareness of the file drawer came early to psi research, that Kanthamani and Broughton’s (1994) studies were not excluded from the relevant meta-analyses (Bem, Palmer, & Broughton 2001; Milton & Wiseman 1999; Storm et al. 2010), and that the observed significant-study rate matches what low average power predicts rather than indicating suppression ^[2]. Irwin’s (2014) survey finding that 40% of respondents reported a failed psi experiment ^[3] speaks to the same charge — that nulls are conducted and acknowledged.

Analysis. The retrieved material documents one exchange: a Skeptical Inquirer critique alleging null-result nullification and selection bias, and a Journal of Parapsychology reply (Baptista & Derakhshani 2014) contesting it on reproducibility, file-drawer, and power grounds. The reply’s claim that psi reproducibility is “similar” to related mainstream fields ^[2] is stated but, in these excerpts, not backed by the mainstream comparison figures. An independent, non-parapsychologist adjudication of that specific comparison is not among the retrieved sources.

For the methodological framing of replication as a topic, ESP-Nexus collects this under Methodology.