UPSC CSE · CSAT · Reading and communication · TOPIC LESSON
Judge whether a passage’s evidence supports its conclusion, and identify what would genuinely change that judgment.
BY THE END, YOU CAN
BEFORE YOU START
Original educational lesson. Work through the examples before revealing the practice answers. This original common-core lesson supports its declared paper scope. Optional subjects, qualifying language papers and the interview are outside this resource.
LEARN THROUGH ORIGINAL PRACTICE
All teaching examples, cases, prompts and model answers are original. Official exam questions and answer keys are not reproduced. This common-core lesson does not imply complete historical paper coverage or an official score.
01 / UNDERSTAND & APPLY
An evaluation begins with the exact conclusion. A descriptive claim says what is happening or how common something is. A causal claim says one factor produces or changes another. A prescriptive claim says what someone should do. These claims require different bridges: an observed association may support a description without establishing a cause, and even reliable causal evidence does not alone decide every value tradeoff in a recommendation.
Distinguish validity from empirical credibility. An inference can follow deductively from premises that turn out to be false; a well-supported empirical conclusion can remain uncertain without being irrational. The relevant question is not whether evidence proves everything beyond doubt, but whether it supports the strength and scope of this conclusion.
Name the bridge in ordinary language: “Respondents must represent the population”, “The comparison must isolate the intervention”, or “The measure must reflect the outcome we care about”. Once the bridge is visible, useful strengthening and weakening information becomes easier to recognize.
ORIGINAL PRACTICE · 1
The sampling frame excludes nonmembers, whose views may differ.
Membership could be related to approval of the club. A unanimous member response still does not establish a resident-wide opinion. This criticizes the generalization; it does not prove that most residents oppose the club.
02 / UNDERSTAND & APPLY
The population is the group a conclusion concerns. The sampling frame is the group from which observations could be selected. Respondents are those who actually provide data. A gap at either stage can matter: an app-based frame misses people who cannot use the app, and optional responses may disproportionately attract especially pleased or displeased users.
A large sample reduces some random variability under appropriate sampling conditions, but size alone does not repair systematic exclusion. Repeating the same biased recruitment method a thousand times can yield a very precise description of the wrong group. Nonresponse matters when nonrespondents differ in ways relevant to the question.
Do not label every small sample worthless. A small observation can reveal a possible problem or generate a hypothesis. The fault is claiming a population proportion or universal conclusion without a suitable basis. Keep the evidence’s use proportional to its design.
ORIGINAL WORKED EXAMPLE
Original passage: A station manager placed an optional satisfaction survey inside the station’s travel app. During a week,100 users chose to answer and 90 described the service as satisfactory. The station normally serves 2000 daily users, including people who buy tickets at a counter and do not use the app. A report announced that 90% of all daily station users were satisfied. The answers are useful accounts of the respondents’ experience. They may identify features worth keeping and problems worth investigating. But the report has moved from a voluntary app response to a claim about a larger population. It supplies no method for ensuring that app users resemble counter users or that people who chose to answer resemble people who ignored the survey. Counting more responses from the same recruitment channel might leave those differences intact. A careful evaluation should question the population estimate without inventing the opposite result or dismissing every respondent’s experience. Is the 90% population conclusion warranted?
Answer No. Ninety percent of respondents reported satisfaction; the satisfaction proportion among all daily users remains unestablished.
03 / UNDERSTAND & APPLY
A variable can move with an outcome without causing it. A common cause can affect both, the outcome can influence the supposed cause, or people can select themselves into the intervention. For example, highly motivated participants may both attend a workshop and improve their performance; their motivation is a possible confounder.
Before-and-after comparisons also face other changes over time. A comparison group can help if it shares relevant background conditions but differs in exposure to the intervention. Merely naming another group is insufficient when its rainfall, starting condition, measurement method or concurrent treatment differs in ways that affect the outcome.
For reverse causation, ask whether the outcome might prompt the intervention. More repairs may appear alongside worse equipment because damaged equipment attracts repair work, not because repair work damages equipment. A causal evaluation should consider the direction rather than silently assuming it.
ORIGINAL WORKED EXAMPLE
Original passage: Residents of a resurfaced street report fewer puddles after the work. A short article attributes the improvement entirely to the new surface. During the same period, however, rainfall was lower than in the weeks before the work, and the drainage crew cleared blocked outlets along the street. Both changes could affect the amount of standing water. The new surface may have contributed. Yet the observed improvement alone cannot show how much each factor mattered. Comparing the street with another street would help only if the comparison supplied relevant information about rainfall, drainage work and earlier puddling. A dry street in a different part of the region would not automatically provide a useful control. The aim is to distinguish explanations, not merely to collect another favourable observation. A careful conclusion would retain the improvement as a reported outcome while treating its exact causal allocation as unresolved by the facts supplied. What weakens the sole-cause claim, and what comparison would help?
Answer The concurrent rainfall and drainage changes weaken exclusive attribution. A comparison matched on relevant conditions would be more informative than a bare second observation.
ORIGINAL PRACTICE · 1
The comparable-street finding is more directly relevant.
It helps distinguish the lamp intervention from a general time trend, while stable patrol arrangements address one alternative. It does not remove every possible confounder or guarantee causation. Liking the colour may matter for preference but does not directly explain theft changes.
04 / UNDERSTAND & APPLY
A count measures total occurrences. A rate relates occurrences to an exposure or population. More complaints can accompany a lower complaint rate if visits increase faster. Always specify the denominator and period: complaints per visit differs from complaints per person and from the proportion of people who complain.
Measurement can also change. A new reporting channel may reveal previously unrecorded problems; a stricter definition can raise the recorded count without an equivalent change in the underlying condition. Conversely, an inaccessible complaint system can suppress reports while problems persist. Ask whether the measure stayed comparable.
Rates are not automatically the final answer. Total workload matters for staffing, and a low rate can coexist with many harmed people in a large system. Choose the measure that matches the conclusion instead of replacing one universal shortcut with another.
ORIGINAL WORKED EXAMPLE
Original passage: A service desk recorded 20 complaints during 1000 visits in one month and 30 complaints during 3000 visits in the next. The manager said the service must have worsened because the complaint count increased. Another employee replied that it must have improved because the complaint rate fell. The figures support parts of both descriptions, but not either confident judgment about underlying quality by themselves. The second month involved more than twice as many visits, so the count and rate need not move in the same direction. It is also unknown whether complaint reporting became easier, whether the visits involved different services, or whether complaint seriousness changed. The desk must handle 30 recorded complaints rather than 20, which matters for workload. Evaluating service quality requires more than that workload count, just as it requires more than a single rate detached from context. Calculate both changes before deciding what the evidence actually establishes.
Answer Recorded complaints rose 50%, while complaints per visit fell from 2% to 1%. “Must have worsened” and “must have improved” both exceed the supplied quality evidence.
ORIGINAL PRACTICE · 1
No. The number of services, delay per service and service mix need checking.
A system could run half as many trains with the same delay per train and obtain a lower total. The total remains useful for some questions, but does not by itself identify punctuality per service.
05 / UNDERSTAND & APPLY
A strengthening fact makes the conclusion better supported relative to alternatives; it need not prove it. A weakening fact makes a relevant bridge less secure; it need not establish the opposite conclusion. Judge each option by its relation to the exact claim, not by whether it sounds generally positive or negative.
If a claim says a route is accessible to every user in a defined group, one valid counterexample can defeat that universal. If a claim says a design tends to help under certain conditions, one failure outside those conditions may do little. Match the evidence’s domain to the claim before assessing its impact.
A relevant question often discriminates between rival explanations. If improvement may reflect a reporting change, whether the reporting method changed is more useful than whether staff enjoyed the launch event. Distinguish relevant uncertainty from information that is merely interesting.
ORIGINAL PRACTICE · 1
The universal claim is false under the supplied facts.
The counterexample falls inside the exact set and condition asserted. It does not tell us how common inaccessible points are across the system, nor does it establish that every route is inaccessible.
06 / UNDERSTAND & APPLY
An analogy transfers a conclusion from one case to another based on shared structure. Its strength depends on similarities relevant to the proposed mechanism. Two buildings can look alike while serving radically different needs; two visually different services may share an important scheduling problem.
Identify which feature made the first arrangement work and whether it is present in the second. Then examine relevant differences such as urgency, access, scale, demand variability and required expertise. An analogy can suggest a trial or question without justifying immediate universal adoption.
ORIGINAL WORKED EXAMPLE
Original passage: A small archive introduced appointment-only access and found it easier to retrieve boxes before visitors arrived. A proposal recommends the same arrangement for a busy walk-in emergency repair centre because both services occupy buildings of similar size. The archive’s users usually know in advance which collections they want, while many repair-centre visitors arrive with faults that suddenly prevent essential equipment from working. The proposal identifies a real administrative attraction: predictable arrivals can help allocate staff. Yet building size does not establish that users can safely postpone visits or predict their need. An appointment system might be useful for routine repairs while an urgent walk-in route remains necessary. The archive’s experience therefore raises a design possibility, rather than settling the entire access policy for the repair centre. The comparison should focus on how demand arises and what delay costs users, not on an easily noticed physical resemblance. What limits the analogy?
Answer The urgency and predictability differences limit transfer. Similar building size is superficial relative to the scheduling mechanism.
07 / UNDERSTAND & APPLY
Absence of evidence is not automatically evidence of absence, but it can be informative when a search had a strong opportunity to detect the thing if it were present. Ask about sensitivity, coverage and conditions. A careful inspection for visible large cracks differs from a casual glance in poor light; neither automatically establishes the absence of microscopic damage.
This is one qualified principle, not a blanket ban on negative evidence. An exhaustive, accurate inventory of a sealed container can rule out an unlisted item within that model. Tracks missing after heavy rain are much less informative about whether an animal passed earlier. The strength of the conclusion must match the search.
Keep factual uncertainty separate from values. A policy might demonstrably save average time yet burden a group without access to the required device. The time finding does not settle whether the burden is acceptable. A reasoned recommendation explains both the empirical expectation and the fairness priority rather than presenting the latter as a measured fact.
ORIGINAL WORKED EXAMPLE
Original passage: An inspection team checked the entire visible surface of a panel using a method that reliably detects cracks wider than 2 mm under the stated conditions. It recorded none. A summary declared the panel entirely free of cracks. The inspection was careful within its defined range, but the method’s detection threshold matters to the conclusion. A narrower crack could remain undetected even if the team followed the procedure correctly. At the same time, it would be misleading to say that the inspection provides no information at all. Given reliable detection and full stated coverage, the result is evidence against cracks wider than 2 mm on that surface. It does not automatically address concealed surfaces, smaller cracks or future deterioration. Interpreting the negative finding requires preserving those distinctions rather than either exaggerating it into total assurance or dismissing a useful bounded check. What conclusion is justified?
Answer The finding supports absence of detectable cracks wider than 2 mm on the inspected surface under the stated conditions. It does not establish that the panel has no cracks of any kind.
ORIGINAL PRACTICE · 1
No. The rain could erase the evidence even if an animal had passed.
Detection conditions are poor for the past event. The counterexample is an animal crossing before its tracks wash away. This does not establish that an animal did pass; it limits what the absence supports.
ORIGINAL PRACTICE · 2
No object in that sealed box contains copper at the inventory time.
Exhaustiveness and accuracy remove the missed-object alternative within the stipulated model. The conclusion is bounded to that box and time; it does not extend to nearby boxes or subsequent additions.
ORIGINAL PRACTICE · 3
No. The evidence about time does not settle the access and fairness judgment.
A defensible decision must address the excluded group, alternatives such as assistance, and the values used to compare outcomes. One can accept the measured time benefit while rejecting an unsupported claim that it resolves every policy consideration.
08 / UNDERSTAND & APPLY
Use a four-part evaluation: name the conclusion, identify the evidence, expose the vulnerable bridge, and specify information that would discriminate between the leading explanations. End with the strongest conclusion the evidence supports, rather than a vague “more research is needed”.
Do not demand impossibly perfect evidence for every modest claim. A small pilot can show that an arrangement is feasible for some participants, while remaining insufficient for a population-wide prediction. Similarly, criticism of a universal guarantee need not deny every possible benefit. The judgment should change with the claim’s strength.
If choices are offered, explain distractors by their precise failure: irrelevant to the link, outside the domain, compatible with both explanations, or stronger than the data permit. This creates a reusable evaluation method rather than a list of suspicious words.
ORIGINAL PRACTICE · 1
Outreach may help, but the comparison does not isolate its effect or guarantee success.
The pattern is consistent with a useful outreach mechanism, and the report’s cautious wording preserves uncertainty. Unmeasured village differences limit causal attribution. Neither universal failure nor guaranteed success follows from these two cases.
KEEP THE BIG IDEAS
READ WITH CONTEXT
The teaching notes, passages, worked explanations and practice are original. Official references identify the assessment framework and skill scope; they do not endorse this lesson or license official question reproduction.
The 2026 official notice was checked on 12 October 2026 for the broad skill scope. The public landing page did not respond reliably during that check. All teaching, passages, scenarios and solutions here are original educational material; no official questions, answer keys or future-topic guarantees are supplied.
UPSC analytics cover only explicitly reviewed printed papers. Paper-specific parent counts include unresolved classifications; subparts and Essay choices remain separate. Teaching availability is independent of historical observations.