Mission Design, Data Integrity and Responsible Conclusions

After this unit, students should be able to state a relevant mechanistic hypothesis, define measurable observables and units, organize competing nonbiological explanations, calculate an explicitly idealized numerical example, design a contamination-control or instrument-calibration strategy, explain where a synthetic dataset stops being informative, and write a conclusion bounded by the actual evidence. Students should be able to identify links between the molecular scale and the whole-planet scale without confusing a laboratory success with a proven historical sequence.

8.1 Hypothesis testing and quantitative uncertainty

Core explanation. A meaningful hypothesis predicts observations that could turn out differently. Model parameters, noise and nuisance variables should be stated before inspecting the target data whenever feasible. Apparent precision may be misleading if major uncertainty comes from unmodeled alternatives. A concrete measurement plan is: Compare likelihoods under hypotheses with explicitly chosen prior assumptions and sensitivity checks. The interpretation should specify instrument limitations, physical context and which rival models make distinct predictions. The purpose is to test a mechanism, not simply to accumulate correlated observations.

Worked thought experiment. A spectral feature has a small random error but an uncertain calibration baseline. The explanation compatible with biology must identify a causal process, a suitable environment and a route to the observable. A serious alternative is: A statistically precise detection can remain chemically ambiguous. A useful follow-up should compare a paired geological control and a negative instrument blank, while preserving the original location and sample treatment. If both hypotheses predict the same result at current precision, the outcome remains underdetermined.

Quantitative example — Posterior sensitivity with invented prior. Take the fictional inputs P(H)=0.01; likelihood ratio = 2. Under the explicitly stated classroom relation P(H|E) = LR·prior/(LR·prior+1−prior), the result is 0.019802 probability. Teaching prior, not a measured population frequency of life. Vary one measured input, inspect the sensitivity, and distinguish mathematical precision from the error in the physical assumptions. The result is an instructional calculation, not a mission observation or an estimate of the probability that extraterrestrial life exists.

Exercises and discussion. (1) State exactly what the hypothetical case measures. (2) Explain how a statistically precise detection can remain chemically ambiguous. (3) Write a control that distinguishes source chemistry from analytical contamination. (4) Recalculate the example after increasing one input by ten percent and justify units. (5) Identify an independent measurement that would discriminate mechanisms. (6) Write a two-sentence conclusion containing both the result and its major uncertainty.

8.2 Orthogonal measurements and biosignature ladders

Core explanation. Independent methods can connect morphology, chemistry, isotope ratios, environmental context and temporal behavior. Agreement is strongest when the failure mechanisms are different, not merely when multiple plots result from one calibrated dataset. Evidence ladders are practical frameworks for assessing a claim, not automatic certificates of life. A concrete measurement plan is: Design three complementary tests with separately traceable failure modes and controls. The interpretation should specify instrument limitations, physical context and which rival models make distinct predictions. The purpose is to test a mechanism, not simply to accumulate correlated observations.

Worked thought experiment. Microscopy, Raman spectra and mass spectrometry all analyze one exceptionally intriguing grain. The explanation compatible with biology must identify a causal process, a suitable environment and a route to the observable. A serious alternative is: A preparation reagent affecting all aliquots can correlate seemingly separate evidence. A useful follow-up should compare a paired geological control and a negative instrument blank, while preserving the original location and sample treatment. If both hypotheses predict the same result at current precision, the outcome remains underdetermined.

Quantitative example — Posterior sensitivity with invented prior. Take the fictional inputs P(H)=0.01; likelihood ratio = 10. Under the explicitly stated classroom relation P(H|E) = LR·prior/(LR·prior+1−prior), the result is 0.091743 probability. Teaching prior, not a measured population frequency of life. Vary one measured input, inspect the sensitivity, and distinguish mathematical precision from the error in the physical assumptions. The result is an instructional calculation, not a mission observation or an estimate of the probability that extraterrestrial life exists.

Exercises and discussion. (1) State exactly what the hypothetical case measures. (2) Explain how a preparation reagent affecting all aliquots can correlate seemingly separate evidence. (3) Write a control that distinguishes source chemistry from analytical contamination. (4) Recalculate the example after increasing one input by ten percent and justify units. (5) Identify an independent measurement that would discriminate mechanisms. (6) Write a two-sentence conclusion containing both the result and its major uncertainty.

8.3 Planetary protection and ethical exploration

Core explanation. Forward contamination can confound scientific searches or threaten environments, while backward contamination requires risk controls for returned materials. Mission-specific planetary-protection requirements depend on destination, activity and applicable policy. Ethical decisions also include stewardship, transparency and the fair communication of uncertainty. A concrete measurement plan is: List contamination pathways from manufacturing through landing and sample return, and prioritize prevention and verification. The interpretation should specify instrument limitations, physical context and which rival models make distinct predictions. The purpose is to test a mechanism, not simply to accumulate correlated observations.

Worked thought experiment. A mission proposes sampling an accessible suspected plume deposit. The explanation compatible with biology must identify a causal process, a suitable environment and a route to the observable. A serious alternative is: Mission contamination can create a biological-like signal and alter the environment being studied. A useful follow-up should compare a paired geological control and a negative instrument blank, while preserving the original location and sample treatment. If both hypotheses predict the same result at current precision, the outcome remains underdetermined.

Quantitative example — Posterior sensitivity with invented prior. Take the fictional inputs P(H)=0.01; likelihood ratio = 50. Under the explicitly stated classroom relation P(H|E) = LR·prior/(LR·prior+1−prior), the result is 0.33557 probability. Teaching prior, not a measured population frequency of life. Vary one measured input, inspect the sensitivity, and distinguish mathematical precision from the error in the physical assumptions. The result is an instructional calculation, not a mission observation or an estimate of the probability that extraterrestrial life exists.

Exercises and discussion. (1) State exactly what the hypothetical case measures. (2) Explain how mission contamination can create a biological-like signal and alter the environment being studied. (3) Write a control that distinguishes source chemistry from analytical contamination. (4) Recalculate the example after increasing one input by ten percent and justify units. (5) Identify an independent measurement that would discriminate mechanisms. (6) Write a two-sentence conclusion containing both the result and its major uncertainty.

8.4 Reproducible computation and open evidence

Core explanation. Reproducibility requires retaining raw data, units, processing versions, calibration records, metadata and uncertainty assumptions. Code publication does not rescue a study if sampling decisions or instrument context are missing. Synthetic classroom datasets must be labeled so they are never mistaken for observations. A concrete measurement plan is: Write a minimal provenance record and compare two pipelines that generate different conclusions from the same raw file. The interpretation should specify instrument limitations, physical context and which rival models make distinct predictions. The purpose is to test a mechanism, not simply to accumulate correlated observations.

Worked thought experiment. Two teams classify a molecular feature differently after applying distinct smoothing methods. The explanation compatible with biology must identify a causal process, a suitable environment and a route to the observable. A serious alternative is: Analyst degrees of freedom and preprocessing choices can produce false confidence. A useful follow-up should compare a paired geological control and a negative instrument blank, while preserving the original location and sample treatment. If both hypotheses predict the same result at current precision, the outcome remains underdetermined.

Quantitative example — Independent-sample contamination. Take the fictional inputs n=10 samples; hypothetical p=0.01. Under the explicitly stated classroom relation P(at least one)=1−0.99^n, the result is 0.095618 probability. Common-mode contamination violates independence. Vary one measured input, inspect the sensitivity, and distinguish mathematical precision from the error in the physical assumptions. The result is an instructional calculation, not a mission observation or an estimate of the probability that extraterrestrial life exists.

Exercises and discussion. (1) State exactly what the hypothetical case measures. (2) Explain how analyst degrees of freedom and preprocessing choices can produce false confidence. (3) Write a control that distinguishes source chemistry from analytical contamination. (4) Recalculate the example after increasing one input by ten percent and justify units. (5) Identify an independent measurement that would discriminate mechanisms. (6) Write a two-sentence conclusion containing both the result and its major uncertainty.

8.5 Communication, discovery claims and next tests

Core explanation. The public may read evidence for organic chemistry as evidence for alien life. Researchers should explain the measured quantity first, describe unresolved abiotic mechanisms and distinguish a candidate from a confirmed conclusion. Replication, external review and long-term follow-up are integral to serious discoveries. A concrete measurement plan is: Draft both a scientific abstract and a public-facing summary with the same uncertainty statement. The interpretation should specify instrument limitations, physical context and which rival models make distinct predictions. The purpose is to test a mechanism, not simply to accumulate correlated observations.

Worked thought experiment. A press release proposes saying that a mission has discovered biological molecules on another planet. The explanation compatible with biology must identify a causal process, a suitable environment and a route to the observable. A serious alternative is: The phrase can conflate molecules used by organisms with molecules that unambiguously require organisms. A useful follow-up should compare a paired geological control and a negative instrument blank, while preserving the original location and sample treatment. If both hypotheses predict the same result at current precision, the outcome remains underdetermined.

Quantitative example — Independent-sample contamination. Take the fictional inputs n=30 samples; hypothetical p=0.01. Under the explicitly stated classroom relation P(at least one)=1−0.99^n, the result is 0.2603 probability. Common-mode contamination violates independence. Vary one measured input, inspect the sensitivity, and distinguish mathematical precision from the error in the physical assumptions. The result is an instructional calculation, not a mission observation or an estimate of the probability that extraterrestrial life exists.

Exercises and discussion. (1) State exactly what the hypothetical case measures. (2) Explain how the phrase can conflate molecules used by organisms with molecules that unambiguously require organisms. (3) Write a control that distinguishes source chemistry from analytical contamination. (4) Recalculate the example after increasing one input by ten percent and justify units. (5) Identify an independent measurement that would discriminate mechanisms. (6) Write a two-sentence conclusion containing both the result and its major uncertainty.

Case definition. A fictional interdisciplinary team is reviewing a proposed astrobiological target related to mission design, data integrity and responsible conclusions. Its science lead receives eight columns of instrument-like readings, contextual notes and uncertainty ranges. The numbers are simulated and have deliberately incomplete information. The task is to decide whether the available resources should fund another measurement, whether a proposed life-related explanation is currently distinguishable, and what sample-handling records would be required to rule out a contamination event. The team must not silently upgrade a promising habitability assessment into a claim that organisms have been detected.

Research workflow, phase one: reconstruct the observation. Read the unit's synthetic CSV in datasets/unit-08-synthetic.csv. Every record should be interpreted as a fictitious spatial or temporal sampling event with a proxy signal and an uncertainty. Plot the raw proxy against sampling index, retain the baseline and calculate a median and interquartile range. Inspect whether signal varies smoothly with the contextual index or whether a small number of points dominate the result. Check for impossible physical values and missing units before attempting any fitting. If you choose to normalize, retain the original so another reader can reproduce the transformation. A nonzero proxy reading is not itself a biosignature because the proxy is intentionally not tied to a unique biological mechanism.

Research workflow, phase two: compare mechanisms. Construct three possible origins for the pattern: one biology-compatible pathway, a plausible inorganic reaction or transport process, and an instrumental artifact. The biological interpretation must identify how energy, matter and relevant environmental stability could support it. The abiotic candidate should identify specific minerals, radiation conditions or chemical pathways rather than the vague claim that 'nature could do it.' The instrumental candidate should identify how data handling could introduce a trend. For each explanation state one observation it uniquely predicts and one that it shares with the alternatives. Investigate whether a proposed test is truly independent. Repeating the same assay on samples prepared in the same batch might increase precision but not reveal preparation contamination.

Research workflow, phase three: evaluate uncertainty. Distinguish counting noise, calibration drift, incomplete spatial sampling, alteration after deposition and structural uncertainty in the chosen mechanism. The first two may be amenable to repeated measurements; the latter three require better design or a different observable. Report an interval or a plausible range when numerical assumptions permit, but do not assign empirical probabilities to the existence of life from invented classroom inputs. A negative result should explain the measurement detection threshold and the range of biological scenarios that the assay might miss. A positive result should describe alternate mechanisms, including routes that make organics, gases, textures or isotope patterns without living processes.

Research workflow, phase four: propose the next instrument. The team can choose one enhanced measurement, but must justify its effect on discrimination rather than on aesthetic resolution. For example, a spatially resolved mineral–organic association might distinguish transport histories, whereas a high-precision repeated spectrum could still be ambiguous if multiple species overlap. Design calibration standards, positive controls, negative blanks and records of transfer between instruments. Specify which data and code versions will be released so that independent investigators can attempt to reproduce the result. If destructive sample preparation is required, explain what sample fraction is retained and how archived material is linked to the original geological context.

Deliverable. Submit an 800–1,200 word fictional study report consisting of an abstract, instrument and data description, three mechanistic hypotheses, one worked idealized calculation, uncertainty table, proposed discriminating follow-up and public-facing conclusion. The public version must not use 'life discovered' language unless the defined evidence threshold has been independently satisfied; nothing in this synthetic exercise satisfies that standard. A strong report may conclude that the proposed biological interpretation is unresolved. Such a conclusion is valuable if it identifies a feasible measurement that could resolve uncertainty.

1. Explain a distinction between a necessary environmental condition and a life-specific indicator.
2. Describe why an organic molecule or water observation cannot independently demonstrate extraterrestrial life.
3. Identify two ways different sensors can share one source of systematic error.
4. Explain how an abiotic production pathway can change the meaning of a candidate biosignature.
5. Distinguish a detection limit from evidence of absence.
6. Provide a physically defined next test and the result that would falsify your preferred interpretation.

The two sliders in this unit operate on simple equations with explicitly specified parameters. Use them to discover the mathematical behavior of the equations, not to predict whether or where life exists. Change the input through the full displayed range, inspect whether the output responds monotonically and identify where the simplifications become unrealistic. A model that reproduces its own intended equation has passed only a software correctness check. Physical validation against independent observations and literature is a separate scientific activity. Take particular care when the displayed result is labeled probability: its prior odds and conditional assumptions are invented, not estimated from alien biospheres.

A review panel receives three candidate life-detection claims: a suggestive microscope texture, a weak chemical peak and two gases tentatively identified in a remote spectrum. First ensure the claims are phrased as observations, not conclusions. Draft a common evidence ledger showing instrument origin, calibration version, competing abiotic pathway, shared uncertainties, contamination-control results and external replication status. Apply a simple hypothetical likelihood-ratio calculation to show why invented priors cannot become measured probabilities of life. Ask each team for an outcome that would force it to revise its favored explanation. Rank the value of an independent measurement that can break a degeneracy over simply acquiring another similar detection. Write a press statement that conveys the evidence, limitations and next planned tests without treating uncertainty as a weakness or using exceptional language unsupported by the results. Finally propose an independent audit of provenance before any public claim.