IB Theory of Knowledge 2.12.31 IA Prompt 31: How Can We Judge When Evidence Is Adequate? Notes

2.12.31 IA prompt 31: How can we judge when evidence is adequate?

What the prompt asks

1. Your prompt is "How can we judge when evidence is adequate?" Examine what would count as enough evidence for a claim or decision, and why that standard is justified.
2. Adequate evidence need not guarantee a true conclusion. It may justify accepting a claim or acting under uncertainty, while leaving a possibility of error. Explain what enough means in the case you discuss.
3. Identify the evidence, the claim it supports and, where relevant, the decision it informs. Practical purposes can affect standards, though adequacy can also be a question about what we are justified in believing.
4. The exhibition uses three objects linked to one official IA prompt, with a maximum of 950 words across their commentaries. There is no fixed word allocation for each object.
5. Two families of object fit here. Things that were used as evidence for something, and things that record where somebody fixed the line: a certificate, a log sheet, a licence, a printed margin of error.
6. Connect the particular judgement to a question about knowledge. Whether a laboratory identification was correct may matter to your analysis, but explain how the method and comparison support judging its evidence adequate.

Adequacy is a threshold for a purpose, and someone has to set it

1. A threshold is a standard for accepting a claim or taking action. It may be set in advance or emerge through practice, and it should be open to reasons about the evidence, uncertainty and consequences.
2. Some thresholds are published and some live in habit. A regulator's approval criteria are written down; the point at which your parents believe you about where you were is not, which does not make it any less of a threshold.
3. Particle physicists conventionally use five-sigma statistical significance as a discovery threshold. It helps guard against random fluctuations under a specified statistical model; it does not remove the need to check systematic errors or the analysis itself.
4. In England and Wales, the civil standard is generally the balance of probabilities, while a criminal jury must be sure of guilt. Different standards can produce different judgements from similar evidence. Explain the jurisdiction and decision rather than treating legal proof as one universal threshold.
5. People may disagree about the evidence, the standard or both. Identify the disagreement in your case before explaining why one person finds the evidence sufficient and another does not.
6. Where a decision creates risks, ask who sets the standard and who bears the consequences of error. Those may be different people, but do not assume they are in every case.

More is not the standard: relevance, independence and fit outweigh sheer amount

1. Several reports may repeat the same source. Trace their origins before treating them as independent confirmation. A later report can still add value if it checks the source or supplies additional evidence.
2. Relevance depends on the claim. Attendance data directly support claims about attendance; using them as evidence of learning requires an argument about the connection.
3. Ask whether the evidence discriminates between the claim and plausible alternatives. A check that would return the same result either way adds little support for choosing between them.
4. Searching many possible patterns increases opportunities for chance findings. Exploratory results can still be useful, especially when the analysis accounts for that search and later evidence checks the finding.
5. The scope of the claim matters. One well-established instance may support that something can happen, while a general claim about how often it happens needs broader evidence.
6. Apply similar care to your commentary. Add evidence or reasons where they improve the justification of an object, while keeping within the total word limit.

Object bank

1. The results table from your own science investigation, with its trials and context identified. Ask why the number and quality of observations were enough for the conclusion you drew.
2. The patient leaflet from a particular medicine box, including any side-effect frequency categories. Examine the evidence behind the information and its limits; safety information may be updated after authorisation as further evidence emerges.
3. A photograph taken when you read a particular lateral-flow test within its instructed reading window. Use its actual instructions and context to examine what the result supported. Do not keep a used test as an exhibition object or treat a later change in its appearance as evidence of the original result.
4. The Lancet's February 2010 retraction of the 1998 Wakefield paper. The notice cites false statements about referrals and ethics approval, rather than sample size alone. It lets you examine how concerns about a source affect the adequacy of evidence drawn from it.
5. A particular fire-alarm test log in your school, with dates and test details. Investigate what the recorded checks establish and what further inspection or maintenance the system requires. One entry is not proof that every component will work in a fire.
6. A dated newspaper page reporting an election poll and its margin of error. Identify which uncertainty the margin represents and what other sources of error the report discusses.
7. The carton of a moisturiser or supplement in your bathroom that prints the percentage of users who agreed with a claim, with the number of people asked in small type underneath it.

The consequences of error can affect the standard

1. Acting too soon and waiting too long can both cause harm. Judging adequacy may involve weighing these risks as well as the quality of evidence; it need not mean choosing one mistake you would prefer to make.
2. A hearing screening may justify referral for a diagnostic assessment without establishing a diagnosis or treatment. The decisions require different information, and false results can carry costs even at the screening stage.
3. Reversibility may matter too. A short trial of a bus route and a decision to demolish a building call for different kinds of evidence and create very different consequences if wrong.
4. On 13 March 2019 the FAA ordered a temporary grounding of Boeing 737 MAX aircraft operated by US airlines or in US territory. It cited new evidence and refined satellite data after two crashes. The decision illustrates precautionary action while investigation continues, rather than a conclusion drawn merely by counting two events.
5. Examine how interests and exposure to harm affect a proposed threshold. A disagreement may reflect those differences, but it can also concern the interpretation or quality of the evidence.
6. Where your object concerns a decision, specify what it was: a referral, a conviction or a temporary grounding. Then explain why the available evidence was or was not adequate for that purpose.

Linking a test log to adequacy

1. Identify the object and its context, explain its connection to the prompt and justify its contribution. These tasks do not require a fixed order or three-sentence structure.
2. For a school fire-alarm log, identify the particular record, building and period. Use the details it records rather than inventing a weekly routine or named person.
3. Find out what each recorded test checked. Routine alarm checks form part of a wider safety system; avoid assuming that one sounder was tested or that a weekly entry establishes the reliability of the whole system.
4. What makes a limited check adequate for its intended purpose, and what does it leave untested? The answer depends on the check and the wider evidence, which you should investigate.
5. Explain the log's particular contribution, perhaps a documented routine standard compared with a less formal judgement elsewhere. That comparison is optional, and the reason for including the log must fit its actual contents.
6. Constructed illustration for analysis: "This log sheet shows that my school checks its alarms regularly, which is important for safety." The sentence does not yet explain what the checks establish or why their evidence is adequate.
7. Develop the link by explaining the scope and limits of the recorded checks, using the applicable procedure as evidence. A fixed interval alone does not establish adequacy.

Failure modes on this prompt

1. Stopping at whether a conclusion was correct. Explain how the available evidence justified the judgement, including what could reasonably have been known at the time.
2. Repeating one justification. Several objects can concern insufficient evidence while contributing different reasons, contexts or limitations. Make those contributions explicit.
3. Using a famous case without a specific object. A known event can supply a useful document or artefact, and you need not own it. Students in the same TOK class may not use the same objects, so check your particular choice.
4. Rewriting the prompt. Use the official wording exactly as given. Keep the question of judging adequacy in view throughout the commentary.
5. Leaving context implicit. Identify the object and explain the circumstances relevant to the judgement; a reader should not have to know you or your school to understand it.
6. Leaving an object under-explained. Allocate the 950-word maximum according to what each commentary needs. Equal word counts are not required, but all three objects need context, links and justification.