Testing Dowsing, Applied Kinesiology and Magnetometer Apps

Dowsing rods, muscle tests and phone sensors produce different kinds of information. Start with the question each method can answer, then look at blinded comparisons, repeatability and the practical limits of a room measurement.

Key takeaways

• Name the target first. Finding concealed water, comparing muscle responses and recording a magnetic field each require a different test.
• Keep the answer hidden. Randomised targets and concealed allocation help test whether an apparent success survives the removal of clues.
• Record misses as well as hits. A repeatable result needs a complete record and an agreed scoring rule.
• A phone reports magnetic readings. Sensor settings, position, orientation and nearby objects belong alongside the numbers in your notes.
• Match the conclusion to the experiment. A finding about one task carries only as far as the conditions actually tested.

A pair of rods crosses at one spot in a room. A muscle feels easier to push down there, and a phone displays a changing number. It’s tempting to connect all three observations. Before deciding what they share, write down what each observation actually contains. Was the person expecting a response, did the phone turn, and was the proposed target independently known? These are the questions behind a fair comparison.

What the community says

Dowsing traditions give the practitioner’s response a central role. The independent British Dowsing glossary (not affiliated with the British Society of Dowsers) distinguishes physical or site dowsing, map or distance dowsing and information dowsing. This vocabulary describes how practitioners organise their experience and gives a starting point for choosing a specific claim to test. British Dowsing, Glossary of Dowsing Terms D–F.

Applied Kinesiology also includes several different practices. Its professional association describes muscle testing alongside history-taking, posture, movement and other assessments. A demonstration using an arm to answer questions about a room needs to specify which procedure it uses. You can take an experience seriously while asking how to check its interpretation. ICAK-USA’s account.

A fair test starts with a precise claim

Choose a target whose answer can be checked independently. For water detection, that might be whether water is flowing through an agreed concealed arrangement. For a proposed Hartmann or Curry grid, first record where the lines are predicted. Agree what counts as a match before another person repeats the mapping, with the first map concealed. Our evidence method explains why the claim and the test must stay connected.

What blinded dowsing tests found

The Munich barn experiments tested the location of a movable water pipe. Enright’s reanalysis disputed the favourable interpretation published by the original investigators, challenging the statistical treatment and repeatability of the apparent successes, which challenged the statistical treatment and repeatability of the apparent successes. Enright’s account.

The Kassel tests in 1990 used concealed water-flow and object tasks. The organisers reported results compatible with chance. Kassel organisers’ report.

Movement itself can have an ordinary mechanical explanation. Motion-capture research on a hand-held pendulum recorded small finger movements driving larger swings. This ideomotor explanation gives a concrete possibility to investigate when movement feels involuntary. The experiment concerned a pendulum, so the mechanism of a particular rod response still needs examination. Cantergi and colleagues.

Muscle tests depend on the question

A conventional muscle test assesses muscular resistance. Inferring a hidden substance or underground feature from that response adds a separate claim. The target needs to be defined and the inference checked against independently known answers.

One controlled test found chance-level identification. In a double-blind experiment involving 51 people, muscle tests and a grip-strength device identified the designated sealed vial in 80 of 151 comparisons, or 53%. The result was compatible with chance. The participants and people conducting the tests were unaware of the vial contents. Schwartz and colleagues.

A different task produced above-chance results. Two experiments, involving 48 and 20 practitioner–participant pairs, reported above-chance muscle-test results for spoken true and false statements. The person speaking knew the answer, while the practitioner was blinded. The authors explicitly limited transfer to other applications. Hidden water and room grids would require their own validation. Jensen and colleagues.

Record the procedure as carefully as the response. Which muscle was used, who applied pressure, how was the person positioned, and when did testing stop? These details let someone reproduce the task. Personal observations also benefit from recording your first impression before seeing a map or interpretation. The notebook method.

What a phone magnetometer measures

The sensor reports the magnetic field at the phone. Android documents three components, aligned with the handset’s axes, in microteslas, written µT. Turning the phone at a fixed spot changes those components even in a steady field. Calibration also matters, and a sensor labelled uncalibrated may still include factory and temperature corrections. Android sensor documentation.

Check whether the app displays individual components or a total magnitude calculated from them. Keep the recorded quantity the same across visits. The mathematical total combines the components, while an individual component depends on how the phone is oriented.

Device behaviour can enter the trace. Tests of several phones found model-dependent glitches and baseline shifts, including changes associated with charging. A step in a graph therefore needs comparison with the recording conditions before it can be assigned to the room. Odenwald’s smartphone study.

A repeatable local change is a starting observation. Check whether it follows a nearby object or stays at a fixed position. Establishing its source requires further investigation. Earth’s field also varies over time; whether a phone can resolve such variations is a separate question, covered in our magnetic-field overview.

Keep a room comparison repeatable

Use this as an observation exercise. Start with a phone you already have and an app that identifies its sensor, units and recorded quantity. Choose positions before inspecting the readings. A simple plan makes it easier to revisit the same places and compare like with like.

1. Fix the geometry. Mark each position, record height and handset orientation, and use the same support.
2. Record the setup. Note the phone, app, calibration state, nearby objects and whether charging is connected.
3. Repeat the route. Keep the setup consistent and retain the complete readings, including uneventful visits.
4. Change one condition. Repeat after moving an easily movable nearby object, then compare the records.

Let the result keep its limits

Separate a room reading from a personal experience. If you also note how you slept, record that first and compare it with the measurements later. Record light, noise and routine on the same nights, and treat any match between sleep notes and readings as an observation to test, not as a cause. Keep the observations alongside the wider questions in our bedroom reading.

FAQ

Why do rods or pendulums move when I feel still?
Small hand movements may occur without a clear sense of directing them. Pendulum experiments have recorded this mechanism. A movement can feel convincing while its interpretation remains open. Checking a concealed target helps evaluate what information the movement supplies.

Does one successful blind test settle the question?
A successful test gives a reason to repeat the same procedure with fresh targets. Agree the scoring rule and number of attempts beforehand, preserve every result and check whether independent repeats retain the performance.

Can an app identify a water vein?
A magnetometer app records a magnetic quantity. Assigning a change to groundwater requires a validated relationship between that reading and the proposed target. None of the sources cited here describes a method that links a phone’s magnetic reading to groundwater.

Can I compare readings from two phones?
You can record both, but differences may reflect their sensors, calibration and placement. Compare repeated changes within each phone under the same conditions. A useful comparison needs the models, apps, units and recording setup alongside the results.

Glossary

• Blinding. Keeping the target’s identity hidden from people whose knowledge could influence a test.
• Randomisation. Assigning targets or trial order using chance before the responses are collected.
• Ideomotor movement. Movement linked to an idea or expectation that can occur without conscious awareness of directing it.
• Magnetometer. A sensor that measures a magnetic field.
• Microtesla. A unit of magnetic flux density, written µT.
• Calibration. Comparing an instrument’s output with a reference to establish their measurement relationship.

Related on SolarHealth

• The notebook method
• Sleeping north or south
• Kp and Schumann measurements

References

• Enright J. T. (1995). Water dowsing. The Scheunen experiments. Naturwissenschaften 82, 360–369. Original analysis. See also the author’s 1999 account of the Munich experiments.
• König R., Moll J. and Sarma A. (1991). Wünschelruten-Test in Kassel. Skeptiker 1/1991, 4–10.
• Leitgeb N. and Lukas R. (2008). Should hospitals protect from geopathogenic zones? Wiener Medizinische Wochenschrift 158, 42–48.
• Cantergi D., Awasthi B. and Friedman J. (2021). Moving objects by imagination? Amount of finger movement and pendulum length determine success in the Chevreul pendulum illusion. Human Movement Science 80, 102879. Original study.
• Schwartz S. A. et al. (2014). A double-blind, randomized study to assess the validity of applied kinesiology (AK) as a diagnostic tool and as a nonlocal proximity effect. Explore 10(2), 99–108. Original study.
• Jensen A. M., Stevens R. J. and Burls A. J. (2016). Estimating the accuracy of muscle response testing. Two randomised-order blinded studies. BMC Complementary and Alternative Medicine 16, 492. Original study.
• Android Developers. Position sensors. Technical documentation.
• Odenwald S. F. (2022, first published online 2021). Can smartphones detect geomagnetic storms? Space Weather 20, e2020SW002669. Original study.
• British Dowsing. Glossary of dowsing terms D–F. Community glossary maintained by Grahame Gardner.
• International College of Applied Kinesiology USA. AK FAQs. Practitioner organisation’s account.