The Emotion and Motivation Laboratory is based in the Ludwig Maximilian University of Munich, under the direction of psychology professor Markus Maier. The lab carried out successful replications of Daryl Bem’s retroactive facilitation experiments in 2014 and since when it has conducted other parapsychological research, mainly in relation to psychokinesis.
- Maier’s laboratory initially reported successful replications of Bem-style retroactive avoidance effects but later also published a preregistered multi-lab failure, treating both outcomes as theoretically informative.
- The lab has focused on micro-psychokinesis and observer effects, especially oscillatory or temporal-change patterns that may not appear in ordinary endpoint analyses.
- The work of Moritz Dechamps and Marissa Jakob show how the lab links psi questions with broader models of unconscious needs, subjectivity and the relation between mind and matter.
Personnel
Markus Maier
The psychology research of Markus Maier centres on understanding unconscious emotional and motivational processes and their impact on behaviour planning and decision making. Within this broad framework he is currently working on experimental tests of the Pauli-Jung model (PJM), a theory relating the duality of mind and matter to a common ground, the unus mundus. This dual-aspect monism theory provides an alternative to purely physicalist models of mind matter interactions. Mathematical formulations of the PJM are the Generalized Quantum Theory and the Model of Pragmatic Information, which both play a major role in his research.
Moritz Dechamps
Moritz Dechamps was awarded a PhD in 2019 with a thesis entitled: Mind Matter Interactions and Their Reproducibility.1Dechamps (2019): https://edoc.ub.uni-muenchen.de/25403/ Together with Maier, Dechamps is investigating quantum-based models of consciousness and the relationship between mind, time and matter, which broadly overlaps with the micro-PK paradigm in parapsychology.
Marissa Jakob
Marissa Jakob was a PhD student (awarded her doctorate in 2025) who examined the role of unconscious needs on PK performance.
Research
Retroactive Avoidance – Feeling the Future Again
In 2014, Maier reported replications of Daryl Bem’s implicit precognition experiments.2Maier, Buechner et al. (2014). Seven experiments tested whether negative stimuli presented in the future could result in an unconscious avoidance reaction in the present. Most experiments found evidence of retroactive influences, supporting Bem’s findings. Across the seven studies participants unconsciously avoided negative future outcomes, with a mean effect size similar to that found by other researchers. The experimenters discuss these findings with regard to quantum models of the mind.
Multi-Lab Replication
In a 2020 publication, Maier and colleagues reported a preregistered multi-lab replication of Maier’s 2014 experiment 4.3Maier, Buechner et al. (2020). The study started in 2013 and ended in 2019. The same design was used as previously, including a sophisticated quantum-based random number generator. The experiment gave evidence for the null hypothesis, an absence of any precognitive avoidance effect (Bayes Factor = 4.38 moderate level evidence).
Given the failure of this well-conducted replication, the validity of the original findings is questioned. Temporal change analyses performed on a post-hoc basis provided significant evidence for non-random oscillations of the retroactive avoidance effect across time, which is in line with the propositions derived from the Model of Pragmatic information. These exploratory analyses of temporal effects in the data, performed on a post-hoc basis, might guide future research in this area. Specifically, the temporal change analysis might help to distinguish false-positive results from true decline effects that otherwise appear phenomenologically identical.
Micro-Psychokinesis – Online Study
In a large study reported in 2018, Maier and Dechamps described testing thousands of participants with the aim of providing clear-cut evidence for or against the existence of micro-PK.4Maier & Dechamps (2018); Maier, Dechamps & Plitsch (2018). An online experiment with 12,571 participants was conducted. Participants first listened to relaxing music in order to induce a positive mood before being presented with a negative or positive image, the sequence determined by a random number generator. It was predicted that more positive pictures would be presented than by chance. A Bayesian analysis (a statistical method that takes prior probabilities into account) carried out by Maier, Dechamps and Pflitsch revealed strong evidence for the null hypothesis H0 (BF01 = 10.07), indicating an absence of micro-PK in these data. However, closer inspection of the temporal change of the effect (how the effect modulated over time) suggested a non-random oscillative structure similar to that seen in the smoker study. Comparison with simulated random data revealed a marginally higher frequency in the experimental portion.
Smoker Study
In 2019, Maier and Dechamps published the results of three studies that used smoking-related stimuli to elicit psychokinetic (PK) responses among 500 smokers and non-smokers.5Dechamps & Maier (2019). In the first study, with 297 participants, they found a strong initial PK effect followed by a strong decline, as predicted by some theories that connect parapsychology to physics. A recovery in PK performance subsequently occurred during the second study, giving an oscillatory effect across all three studies and all 500 participants.
Significantly, the authors predicted this oscillatory pattern from previous experimental results and theoretical development. They explain the pattern in entropic terms: PK ordering results in a loss of entropy which must be balanced later by a fall in PK performance. A recovery in PK scoring then occurs as the system allows for order to be injected again.
In the third study, carried out with 203 participants, Dechamps and Maier predicted not only the continuation of the oscillatory pattern, as described above, but also an approximate PK scoring value based on the oscillation curve. Both predictions were confirmed; however, a comparison with 10,000 simulated random data sets for three measures – an area-under-the-curve analysis, a local maximum fit test, and an endpoint fit test – showed non-significant results.
Further exploratory analyses – including analysis of the highest-reached Bayes Factor (BF) over the course of the experiments, the overall orientation of the BF curve, and its transformation into oscillatory components via a Fourier analysis – showed significant differences between experimental and simulated data. This supports the approach of analysing the temporal development of the PK effect rather than just the terminal value – an approach which, Dechamps argues, more parapsychology programmes should adopt.
Building on this foundation, subsequent research has explored correlational effects between experimental and control data. A reanalysis of the Subliminal Primes series revealed a correlation of 0.032 across 12,254 trials, producing a Bayes Factor of 36.46 indicating very strong evidence of an effect. However, when a preregistered fourth experiment attempted to replicate this correlational effect, the correlation disappeared after correcting for erroneous coding, though limited evidence for experimenter psi was suggested through Change of Evidence measures.6Maier & Dechamps (2022).
The laboratory has also investigated the impact of reducing objectivity on psi outcomes through research with Thomas Rabeyron from the University of Lorraine. In trials where data were recorded subjectively (participants recalled their scores) versus objectively (automatic data storage), strong evidence for PK effects emerged only in subjective trials (Bayes Factor: 17.14), while objective trials showed no effect. A fortuitous data preservation error allowed validation of these findings, supporting the concept of a hybrid ‘sobjective’ reality composed of both objective and subjective aspects.7Maier et al. (2022).
Subliminal Primes
A report submitted in December 2020 by Dechamps, Maier, Pflitsch and Duggan describes the influence of subliminal priming on PK expression.8Dechamps et al. (2021). Four experiments are detailed. In the first experiment, 4,092 participants, recruited online, were presented with 40 positive or negative images, the sequence determined by a quantum random number generator. For experimental primes, before image presentation, a series of three subliminal images are flashed to the subject. The first image is a 50:50 jumbled composite of an equivalent positive and negative picture (for example, an aggressive dog and a polite dog), the second image is a 60:40 composite, and the third image is a 70:30 composite. In this way the positive or negative picture becomes more dominant during the priming sequence and should be more strongly activated in the subject’s unconscious mind. For control primes, only a 50:50 image was delivered before the randomly determined positive or negative picture. Results were strongly in favour of a micro-PK effect caused by subliminal priming on picture selection (Bayes Factor 13.35). Control primes showed no PK influence (Bayes Factor 4.19).
Experiment 2 was an exact replication of Experiment 1 with 2,098 participants. Here, results were well within chance levels. In a post hoc analysis, data from experiment 1 and 2 were combined to examine temporal evolutionary changes that have been found previously. All three temporal measures fell within the top 3% of control data generated from 10,000 simulations, giving strong evidence for anomalous structure across the first two studies combined.
Experiment 3 gave null results only. Experiment 4 gave null results for the experimental condition but interestingly, the control condition was significant (p = 0.01), suggesting a possible displacement effect from experimental to control condition. None of the temporal measures for experiment 3 and 4 were significant. It was concluded that the initially strong results in experiment 1 were not replicated in this series. Given the methodological rigour of the first experiment, the authors suggest the ubiquitous decline effect is responsible for the lack of replication and draw upon emerging theories overlapping parapsychology and physics.
Unconscious Needs
In a report published in 2020, Jakob, Dechamps and Maier report on the effect of subconscious motivational drives on the output of a quantum random number generator (QRNG).9Jakob et al. (2020). Three core psychological needs (attachment, self-esteem protection and control) were measured in 318 participants. Subjects were presented with visual stimuli designed to trigger each of these three needs, the stimuli sequence being determined by the QRNG. Attachment and control needs stimuli produced chance deviations, but self-esteem protection stimuli were significantly deviated by a subset of participants predicted in advance to achieve this (p = 0.02). This group also produced an oscillation over time, reaching very high deviations from chance and then declining.
The second experiment followed this group to determine whether the oscillation effect continued for the self-esteem stimuli. This proved to be the case, reaching a Bayes Factor of 10.72 (strong evidence) before declining to a Bayes Factor of 2.24, the level of anecdotal evidence but significant when measured with regular frequentist statistics (p = 0.03). Post hoc comparisons of three temporal measures with 10,000 simulations indicated that the data for the selected group significantly deviated from chance fluctuations.
It was concluded that subconscious beliefs can manifest through volatile yet statistically detectable deviations from quantum randomness. Subsequent preregistered research has investigated personality-mediated observer influences, revealing strong evidence for micro-PK effects only in individuals with high dependent personality traits (BF₁₀=10.41, p<.001), while avoidant and obsessive-compulsive groups showed null results. However, replication attempts using Higher-Level Analytical Strategies completely failed to reproduce these findings, aligning with theoretical predictions that psi effects are inherently non-replicable.10Jakob et al. (2024).
Complementarity Studies: Colour and Intelligence
The lab has pursued a series of studies examining what Maier terms ‘macroscopic complementarity’ — the hypothesis, derived from Generalized Quantum Theory (GQT), that the subjective and the objective aspects of a phenomenon form complementary, non-commuting subsystems linked by acausal correlations rather than by causal transmission. The specific conjecture under test is non-commutability: if the two aspects are complementary, then performing a measurement on one of them should alter the state of the whole system, including the values obtained from the other. The erasure manipulation is the operationalisation of that conjecture. Whether the objective parameters of a stimulus are permanently stored or deleted is a variation in the measurement of the objective subsystem, and the prediction is that this variation will appear in the subjective evaluation of the same stimulus, even though participants are never informed of the manipulation and could not detect it by any ordinary route.
In a 2025 paper in the Journal of Anomalistics, Maier and Dechamps reported that whether or not objective colour parameters (hue and lightness) were stored or deleted by the experimenter systematically affected participants’ subjective likeability ratings of those colours, across four studies. The design deliberately retained a confound in the assignment of colours to conditions, which was ruled out afterwards by permutation analyses on the combined data; a second dependent variable, the variability of brightness judgements, produced initially promising results that did not replicate and were reported by the authors as false positives. A similar series of studies examining intelligence self-assessments initially showed promising results but ultimately failed to replicate, following the characteristic effect-and-decline pattern predicted by the Model of Pragmatic Information.11Maier & Dechamps (2025); Maier et al. (2024).
Colour Erasure
If the non-commutability conjecture is the hypothesis under test, the NT (non-transmission) axiom sets the terms on which it may be tested. On the authors’ reading it forbids a local causal use of non-local correlations, so that a strictly causal experimental test should yield an initial effect followed by a rapid decline. The two colour series stand on opposite sides of this restriction, and the contrast is the point of the second paper rather than an incidental change of method. In the 2025 studies a confound was deliberately retained in the assignment of stimuli – limited colour sets fixed to conditions in advance – so that no single study permitted an unambiguous causal reading; the effect was instead established indirectly across studies by ruling out the alternative explanation, and no decline was predicted or found. Maier, Vogel, Storch and Dechamps, reporting in the Zeitschrift für Anomalistik, then removed the confound.12Maier, Vogel, Storch, & Deschamps (2026). Colours were assigned by randomisation at the participant level, which made a strict causal test possible for the first time and turned the study into a direct test of the NT axiom’s prediction. That prediction was borne out. Strong evidence for the erasure effect accumulated (BF₁₀ = 39.77 at n = 2,946) and was then followed by a clear decline within the same study, ending at BF₀₁ = 6.16 at N = 6,448. A permutation analysis found p < .03 for observing such a pattern by chance. The confounded and the unconfounded series therefore test complementary predictions of the same framework rather than replicating one another, and a reader should not take the decline in the second as failing to support the first.
Three commentaries appeared alongside the two 2026 papers. Thomas Rabeyron, in ‘Beyond the Principle of Elusiveness?’, situates the work against the limits of the Rhinean paradigm, particularly the psi paradox and the principle of elusiveness, and examines what he calls the erasure-confound paradigm as the combination of two innovations: the shift from a volitional task to a perceptual evaluation, and the deliberate introduction of a confound intended to keep the effect in place. Ian Tierney, in ‘Macroscopic Complementarity – Questions and a Conjecture’, accepts that the simplicity of the method and the clarity of the analysis leave little room for conventional alternatives, but sets the complementarity reading against an experimenter-psi reading and proposes a conjecture that might separate them. Harald Walach, in ‘Chasing the Wind: Empirical Signatures of Generalized Entanglement’, places the findings in the context of nearly two decades of unsuccessful attempts to test the Generalized Entanglement Hypothesis directly. The concern the three share is whether the erasure-confound paradigm can genuinely escape the decline that the NT axiom implies.
Maier and Dechamps replied in a joint response. They argue that the NT axiom is best read as an epistemic restriction that arises whenever causal-deterministic documentation is applied to phenomena containing an autonomous subjective component, and that the confounded design therefore achieves only a ‘reduced-objective’ documentation at the level of the single study, which decouples the accumulation of evidence from the arrow of time. They then put the decline prediction to an empirical test. Three previously unpublished colour-erasure-with-confound studies are reported which, together with the four studies of 2025, constitute the complete record of seven studies in this paradigm with no file drawer. The effect appeared in the predicted direction in every one of the seven (sign test p = .008); combined permutation analyses across three independently drawn 36-colour sets ruled out the colour-selection-bias confound as a joint explanation; a meta-regression on the seven effect sizes returned a downward slope that was not significant; and the seventh study still showed a clear effect (BF₁₀ = 3,679.73, d = .30). The authors read this as a stable, reduced-objectively documented acausal correlation rather than the attrition that the NT axiom would predict under direct replication.13Maier & Dechamps (2026).
Food Image Erasure
In the same issue of the Journal of Anomalistics, Maier, Vogel and Dechamps reported a replication in a different stimulus domain.14Maier, Vogel, & Dechamps (2026). Participants rated the tastiness of food photographs while objective parameters — caloric values and stimulus IDs – were stored or permanently deleted. Deliberate stimulus-set confounding was again used. Four preregistered studies were run, total N = 647. Study 1 found BF₁₀ = 1,309.22, d = 0.66. Studies 2a, 2b and 2c replicated with independent stimulus sets; BF₁₀ values ranged from 10.40 to 26 × 10¹⁵, d values from 0.26 to 0.50. A stratified permutation test (p = .002) and multilevel model (β = 4.72, p = .011) ruled out biased stimulus assignment. Sensitivity analyses agreed. The food series thus follows the 2025 colour design rather than the 2026 one: the confound is retained, the effect is documented across studies, and no decline is predicted or observed.
Biased Baselines
The LMU team have investigated the influence of subject beliefs on outcomes by manipulating baseline probabilities of quantum generated random events. Participants were assigned to either a ‘lucky’ condition with a 60% baseline win probability or an ‘unlucky’ condition with a 40% baseline win probability. In both conditions, a micro-PK biasing effect towards 50% was expected. The LMU team found that ‘lucky’ participants (N = 801) provided strong evidence for a shift toward 50 % (11.80 where 12.00 was predicted by chance, p = .004). The unlucky group (N = 725) showed null results. The lucky-group finding suggests expectation alignment can support micro-PK and limit decline effects; emotional detachment may explain why the unlucky-group produced null findings only.
Doctoral Research: Onkes
Mia Onkes has been a doctoral student since 2022, supervised by Markus Maier and Prof Dr Vanessa Büchner, and submitted her thesis in January 2026, kindly sharing it with the author.15Onkes (2026). The work examined whether symbolic meaning modulates psychophysical correlations between mental states and physical events. Three preregistered studies were run, total N = 3,352. Study 1 (N = 801) did not support the classical micro-PK mean-shift hypothesis (BF₀₁ = 2.32). In the control condition, negative affect correlated with quantum RNG selection of negative stimuli (r = .21, BF₁₀ = 3,400). Under symbolic priming using Omega (Ω) and Psi (Ψ) this coupling was strongly attenuated (r = .06, BF₁₀ = 0.40). Study 2a (N = 457) found the correlation in unvaccinated participants (r = .20, BF₁₀ = 30.46) but not in vaccinated participants (r = -.09, BF₀₁ = 8.26). Study 3 (N = 1,608) generalized the paradigm to inflation-related stimuli and new control symbols, finding moderate evidence for correlation attenuation under ΩΨ (BF₁₂ ≈ 4), below the preregistered threshold of BF > 10.
Doctoral Research: Schweizer
Martin Theodor Schweizer submitted a doctoral dissertation in March 2026, supervised by Maier and Prof Dr Günter Schiepek, and kindly shared with the author.16Schweizer (2026). Three empirical programmes are presented. The first tested whether placebo priming improves cognitive performance; two preregistered studies returned null results. The second tested whether the same priming biases quantum RNG output across two preregistered studies, combined N exceeding 3,500. Main analyses were null (BF₁₀ = 0.30, experimental group, Study 1; BF₁₀ = 1.29, control group). Research Paper 3 (N = 2,000) investigated whether habitual physical activity and nondual awareness moderated quantum RNG forced-choice performance. Confirmatory analyses supported the null (BF₀₁ = 3.904, full sample). Subgroup analyses were also null.
Complete LMU Micro-PK Database
Moritz Dechamps published a publicly accessible meta-analysis in 2025 of the complete LMU Munich micro-PK database, covering all studies from 2016 to 2025.17Dechamps (2025). Every experiment conducted under Maier’s supervision is included regardless of outcome. Overall effect sizes are marginally significant. Temporal dynamics analyses provide more consistent evidence of anomalous structure than simple directional tests. The meta-analysis is maintained as a living record and is updated as new studies are completed. The accompanying manuscript was not accepted at PLOS One, where no handling editor could be found for it; resubmission elsewhere is planned.
The application of artificial intelligence
The artificial-intelligence work at LMU is not primarily psi-related. The lab intends to extend it into this domain in future.
Future Plans
The longer-term aim is to establish ‘sobjective’ reality as the core level at which psychophysical interactions are to be described, that is, the level at which subjective experience and intention and their objective manifestations or counterparts are two aspects of one underlying reality rather than two domains exchanging signals.
Collaboration with other groups
The lab collaborates with the Institut für Grenzgebiete der Psychologie und Psychohygiene (IGPP) in Freiburg, and with Thomas Rabeyron’s group at the Université Lumière Lyon 2 (Centre de Recherches en Psychopathologie et Psychologie Clinique), where Rabeyron is also a junior fellow of the Institut Universitaire de France.
Michael Duggan
Works Cited
Dechamps, M.[C.] (2019). Mind-Matter Interactions and their Reproducibility. [Full text.] Dissertation, Faculty of Psychology and Educational Sciences, Ludwig Maximilian University (LMU), Munich, Germany.
Dechamps, M.C. (2025). Meta-analysis of a complete micro-PK database from the LMU lab: Years 2016–2025. Zenodo.
Dechamps, M.C., & Maier, M. (2019). How smokers change their world and how the world responds: Testing the oscillatory nature of micro-psychokinetic observer effects on addiction-related stimuli. Journal of Scientific Exploration 33, 406-34.
Dechamps, M.C., & Maier, M., Pflitsch, M., & Duggan, M. (2021). Observer dependent biases of quantum randomness: Effect stability and replicability. [Download PDF.] Journal of Anomalous Experience and Cognition 1/1-2, 114-55.
Jakob, M.J., Dechamps, M.C., & Maier, M.A. (2020). You attract what you are: The effect of unconscious needs on micro-psychokinesis. Journal of Parapsychology 84, 227-53.
Jakob, M.J., Dechamps, M.C., & Maier, M.A. (2024). Testing the effects of personality-related beliefs on micro-PK. Journal of Anomalous Experience and Cognition 4/1, 34–59.
Maier, M.A., Buechner V.L., Dechamps M.C., Pflitsch M., & Kurzrock W., Tressoldi, P., Rabeyron, T., Cardeña, E., Marcussen-Clavertz, D., & Martinskovskaja, T. (2020). A preregistered multi-lab replication of Maier et al. (2014, Exp. 4) testing retroactive avoidance. PLOS One 15/8, art. e0238373.
Maier, M.A., Buechner, V.L., Kuhbandner, C., Pflitsch, M., Fernandez Capo, M., & Gamiz, M. (2014). Feeling the future again: Retroactive avoidance of negative stimuli. Journal of Consciousness Studies 21, 9-10.
Maier, M.A., & Dechamps, M.C. (2018). Observer effects on quantum randomness: Testing micro-psychokinetic effects of smokers on addiction-related stimuli. Journal Scientific Exploration 32/2, 265-97.
Maier, M.A., & Dechamps, M.C. (2022). A pre-registered test of a correlational micro-PK effect: Efforts to learn from a failure to replicate. Journal of Scientific Exploration 36/2, 251-63.
Maier, M.A., & Dechamps, M.C. (2025). Macroscopic complementary relation between subjective observations and objective measurements of colors. Journal of Anomalistics 25/1, 15-60.
Maier, M.A., & Dechamps, M.C. (2026). Acausal psychophysical correlations in a causally closed world? Evidence from (un)confounded erasure paradigms and the limits of physicalist methodology. Zeitschrift für Anomalistik 26/1, 120–145.
Maier, M.A., Dechamps, M.C., & Plitsch, M. (2018). Intentional observer effects on quantum randomness: A Bayesian analysis reveals evidence against micro-psychokinesis. Frontiers in Psychology 9, 379-92.
Maier, M.A., Geissler, C., DeMattia, E.C., Vogel, A., & Dechamps, M.C. (2024). Complementarity of subjective and objective realities: An experimental investigation of variations of subjective self-reported intelligence data when objective data are erased or not-erased. Journal of Anomalistics 24/1 15–54.
Maier, M.A., Vogel, A., & Dechamps, M.C. (2026). Macroscopic complementarity between subjective and objective food image assessments. Zeitschrift für Anomalistik 26/1, 37–74.
Maier, M.A., Vogel, A., Storch, J., & Dechamps, M.C. (2026). Non-classical correlation between subjective and objective color observations. Zeitschrift für Anomalistik 26/1, 14–36.
Onkes, M. (2026). Beyond the Binary: Symbolic Modulation of Psychophysical Correlations — Meaning and Affect in a Dual-Aspect Framework. Inaugural-Dissertation, Ludwig-Maximilians-Universität, Munich, Germany.
Schweizer, M.T. (2026). Expectancy and Reality: Perspectives on Mind-Matter-Interactions. Inaugural-Dissertation, Ludwig-Maximilians-Universität, Munich, Germany.