The one idea
Most contradictions are not contradictions. They are two sources answering slightly different questions, and the difference got lost when someone summarised them.
So the useful move is not “which one do I believe?” It is:
What would have to be true for both of these to be correct?
Nine times out of ten there is an answer, and finding it gives you something far more valuable than picking a side: you understand the actual shape of the thing.
When two sources clash, look for what is different about them before deciding one is wrong.
Compare population, measure, method, time period and who funded it. State a conclusion with its conditions attached.
Distinguish conflicting results (different findings on the same question) from non-comparable results (different questions, different populations, or different operationalisations of the same construct).
The five differences to check
When two credible sources disagree, one of these five is almost always the reason. Check them in order — the first two explain most cases.
1. Different population
Who was studied? Remote work research on software engineers tells you very little about remote work for hospital administrators. “Workers” is not a group.
2. Different measure
What exactly was counted? “Productivity” can mean output per hour, tasks closed, revenue per employee, self-reported focus, or hours logged. Two studies using different measures are not in conflict — they are not in conversation.
3. Different method
Survey, observation, controlled trial, or company case study? A self-reported survey and a measured trial routinely produce different answers about the same thing, and neither is lying.
4. Different time period
Was this measured during an unusual period? Data collected during a disruption measures the disruption, not the underlying effect. Anything about work practices from 2020–2021 needs this asterisk.
5. Different interest
Who paid for it, and what result would have suited them? Not a disqualifier — but if a study funded by a vendor finds strongly for that vendor's product, that is worth stating rather than hiding.
Recommendation If you can identify which of these five explains the gap, you no longer have a
contradiction. You have a finding with conditions — which is what a real answer
looks like.
Working through a real conflict
Two studies, opposite conclusions
1 of 6State both claims precisely, in their own words. Not “one says it works and one says it doesn't”. Write the actual claim, with its numbers and its population, as the source stated it. Half of all apparent conflicts dissolve at this step, because the summaries were sloppier than the sources.
Try this
Here are two real-sounding claims:
A: “A trial of 61 companies found no fall in revenue under a four-day week.”
B: “A survey of 500 managers found 60% reported reduced output after shortening hours.”
Before reading on: name three differences that could let both be true.
Your challenge
Level 3 · IndependentPick a genuinely contested question in a field you care about. Some options: does homework improve learning outcomes; do open-plan offices reduce collaboration; does a particular supplement do anything.
Find one credible source on each side. Then produce, in under 300 words:
- Each claim, stated precisely as its source states it.
- Which of the five differences explains the gap.
- A conclusion with its conditions attached.
- One sentence naming what you still do not know.
The fourth point is not a weakness in your answer. It is the part that makes the rest trustworthy.
What people usually get wrong
- Picking the source that agrees with you. Everyone does this unconsciously. The defence is to write down what you expected to find before you search, then notice if you found exactly that.
- Counting sources instead of weighing them. Eight blogs repeating one study do not outvote one careful study.
- Treating “more recent” as “more correct”. Newer is better for prices, rules and tools. For research, a large careful older study can beat a small recent one.
- Splitting the difference. If one source says yes and one says no, the answer is rarely “sort of”. It is usually “yes under these conditions, no under those”.
- Hiding the uncertainty to sound confident. Stating your limits reads as competence. Being caught overstating reads as carelessness, and it is much harder to recover from.
- Asking an AI to settle it. An AI will produce a confident synthesis that may quietly flatten exactly the differences you were trying to find.
How someone experienced does it
Experienced people look for the review before the studies. If a topic has been argued for years, somebody has usually gathered the evidence and weighted it. Finding that first saves days and gives you a much better starting map.
They also state their conclusion at the level of confidence the evidence supports, and they use plain calibrating language: “consistently found”, “mixed”, “one study suggests”, “I could not find evidence either way”. Those phrases are not hedging. They are accuracy, and colleagues learn to trust someone who uses them precisely.
And they name the disagreement in the write-up rather than smoothing it over. “The two main studies conflict, and here is why” is a stronger deliverable than a clean answer that falls apart the moment someone else searches.
When not to use this
Do not run a full comparison when the stakes are low or the answer is not contested. Most questions have a settled answer and you can just take it.
Reserve this for questions where the decision is expensive to reverse, where you will be asked to defend the answer, or where you noticed that confident sources are contradicting each other — which is itself the signal that the question is harder than it looks.
Prove it
Produce a one-page brief on a contested question, structured as: the question, what the strongest evidence on each side actually claims, why they differ, your conclusion with conditions, and what you still do not know.
This is the single most transferable artifact in this module. It is recognisably the same document as a market assessment, a policy note, a technical evaluation and a due-diligence summary.
Keep learning this
Paste this into any AI assistant. It turns the assistant into a tutor that tests you instead of just answering you.
Act as an experienced practitioner who is good at teaching. I have just learned resolving conflicts between credible sources. Assume I am intelligent but relatively new to this — treat me as intermediate level. Work through this in order, and wait for my reply at each step: 1. Ask me 5 questions that test whether I actually understood resolving conflicts between credible sources. Do not reveal the answers yet. 2. After I answer, tell me which parts I got right, which I got wrong, and which I only half-understand. Explain only what I misunderstood — do not re-teach what I already know. 3. Give me one practical challenge based on something I could genuinely encounter at work or in daily life. Do not solve it for me. 4. Evaluate my solution the way an experienced person would judge it, including what a professional would have done differently. 5. Tell me what to learn next, and why that comes next. 6. Give me trustworthy sources for deeper study — prefer official documentation, primary research or standards bodies over blogs and videos. Rules for you: no buzzwords. No motivational filler. Say "I'm not certain" when you are not certain, and tell me which parts of your answer I should verify myself. Clearly separate facts from your recommendations and your opinions.
Become independent at this
Use this when you want a path from where you are to actually good, with checkpoints you can test yourself against.
I want to become independently capable at synthesising conflicting evidence — not permanently dependent on AI, tutorials or step-by-step guides. Design a progression for me with five stages: Beginner, Guided practice, Independent practice, Real-world application, Professional level. For each stage tell me: - what I must know - what I must be able to do without help - the mistakes people make at this stage - one practical challenge - one real project that would prove I reached this stage - one way I can test myself honestly Then tell me the signals that I am ready to move to the next stage, and the signals that I have skipped ahead too early. Keep the theory to the minimum I actually need. Focus on ability I can transfer to situations you and I have not discussed.