🧠 Deductive vs. Inductive Arguments

TypeSupport TypeGoalExample
DeductiveConclusiveGuarantee truth of conclusion“All humans are mortal; Socrates is human → Socrates is mortal.”
InductiveProbabilisticMake conclusion likely“All 100 squirrels seen are grey → next one will likely be grey.”

⚙️ Key Distinctions

  • Deductive Validity: Impossible for premises to be true and conclusion false.
    If true premises, conclusion must be true.
  • Inductive Strength: Unlikely for premises to be true and conclusion false.
    Premises make conclusion probable.

✅ Evaluating Deductive Arguments

QualityMeaning
ValidPremises → Conclusion necessarily
InvalidPossible for premises true, conclusion false
SoundValid + Premises are true

Example:
“All wines are beverages. Chardonnay is a wine → Chardonnay is a beverage.” ✅ Sound.

“Ginger ales are white wines; Vernors is ginger ale → white wine.” ❌ Invalid content – unsound premise.


🔍 Evaluating Inductive Arguments

QualityMeaning
StrongPremises render conclusion probable
WeakPremises unrelated or insufficient
CogentStrong + true premises
UncogentWeak or containing false premise

Example:

  • “Most wines made from grapes. Chardonnay is wine → likely made from grapes.” = Cogent.
  • “Most ginger ales are white wines…” = Strong but false premise → Uncogent.

🧮 Sample Evaluations

  1. Invalid Deduction: “It rains or snows; it rains; so it doesn’t snow.” ❌
  2. Weak Induction: “4 of 5 dentists prefer Crest → most doctors prefer Crest.” ❌ (wrong population)
  3. Strong Induction: “120/135 engineering students favor longer breaks → most engineers do.” ✅
  4. Weak Induction (Overreach): “120/135 engineers → most students on campus.” ❌
  5. Strong + Cogent: Vaccine trial reported with margin of error and representative groups ✅

⚖️ Rules of Evaluation Summary

Argument TypeStrong/Valid WhenWeak/Invalid When
DeductiveConclusion necessarily followsPossible to be false with true premises
InductivePremises make conclusion likelyInsufficient evidence or false assumption

🧩 Terms: Soundness vs. Cogency

DeductiveInductive
Valid + True Premises → SoundStrong + True Premises → Cogent

📦 Examples of Sound vs. Unsound

ArgumentTypeEvaluation
“All wines are beverages; Chardonnay is wine → beverage.”DeductiveSound
“All wines are beverages; Chardonnay is beverage → wine.”DeductiveUnsound (Invalid form)
“Most wines from grapes; Chardonnay is wine → likely from grapes.”InductiveCogent

📊 Inductive Generalizations

Inductive generalizations reason from a sample to a population.

✅ Criteria for a Strong Generalization

  1. Representative Sample — reflects population demographics.
  2. Adequate Size — large enough to minimize error.
  3. Low Margin of Error (M.E.)
  4. Appropriate Confidence Level (CL) — usually 95%.

📉 Statistical Essentials

  • Margin of Error (M.E.)
    • Radius of confidence interval (how far result may vary).
    • Depends on sample size (n) and confidence level.
    • Larger sample → smaller M.E.

Common Formulas

Confidence LevelFormula
99%
95%
90%

Example:
n = 1000, 95% CL → M.E =  (±3%)


📏 Sample Size Guidelines (95% CL, Pop > 15,000)

M.E. ±Sample Size
2%≈ 2000
3%≈ 1000
4%≈ 600

🧮 Interpreting Results – Examples

Example 1

Poll: 2000 students, 51% liberal (±2%, 95% CL)

  • “Majority are liberals.” ❌ Weak (within ME)
  • “49% ± 2% not liberal.” ✅ Strong
  • “Many/some are liberals.” ✅ Strong (safe inference)

Example 2

Poll: 600 students, 54% study daily (±4%)

  • “Most study daily.” ❌ Weak (may dip below 50%)
  • “At least half study daily.” ✅ Strong
  • “1/3 study 2 days/week.” ✅ Strong
  • “Most have high GPA.” ❌ Irrelevant (not linked)

💬 Summary of Strong Inductive Generalization

  • Sample → large & representative
  • Error → small; confidence → high
  • Conclusion matches data (no overreach)

🧩 Evaluating Analogical Arguments

💡 Definition

An analogy compares two things (A & B) that share features (X) and infers they probably share another feature (Y).

Form:

  1. A is like B in respect to X
  2. B has Y
  3. Things with X tend to have Y
  4. ⇒ A likely has Y

✅ Criteria for a Strong Analogy

  1. A truly has property X
  2. B truly has property Y
  3. X is reliably connected to Y
  4. No relevant dissimilarities between A and B

🧮 Examples

Example 1 — Government vs Family Budget

  • Both have credit lines (X)
  • Families → can go bankrupt (Y)
  • ∴ Governments can too
    ✅ Shared traits true, though scale differs → moderate analogy

Example 2 — Thomson’s Violinist (Ethical Analogy)

  • Woman involuntarily pregnant ≈ person involuntarily attached to violinist
  • Person may detach violinist ⇒ abortion justified
    ⚖️ Evaluation:
    • Similarity holds (both sustain another’s life)
    • Disanalogy → mother’s emotional relationship with fetus
    • Outcome → partially strong, morally debatable

Example 3 — Computers & Humans

  • Both solve math problems (X)
  • Humans think (Y)
  • So computers think → ❌ weak analogy (oversimplifies)

Example 4 — Watchmaker Analogy (Paley)

  • Universe & watch = designed (X)
  • Watches → have purpose & maker (Y)
  • ∴ Universe → has maker
    ⚖️ Weak: assumes design implies maker; natural phenomena show pattern w/o intent.

🧾 Diagnosing Analogies

QuestionIf No → Weak Analogy
Do both items really share stated feature X?
Is Y actually present in B?
Is connection between X and Y trustworthy?
Are there relevant disanalogies?

🪶 Summary: Analogical vs. Inductive Generalization

TypeEvidence BaseConclusion TypeCommon Weakness
GeneralizationObserved sampleLikely population trendBiased / small sample
AnalogySimilar casesLikely shared propertyIrrelevant or weak similarity

🧾 Combined Quick Reference

ConceptRule
Deductive ValidityIf premises true → conclusion must be true
Inductive StrengthEvidence raises probability of conclusion
SoundnessValid + true premises
CogencyStrong + true premises
Margin of Error (95%)
Analogical CriteriaShared X, verified Y, reliable link, minimal disanalogy

🧩 Takeaways

  1. Deductive arguments aim for certainty → test structure.
  2. Inductive arguments aim for probability → evaluate evidence quality.
  3. Sound = perfect deduction; Cogent = strong induction.
  4. Generalizations need representative data; Analogies need relevant similarity.
  5. Strong reasoning in logic depends on clarity, precision, and proportional conclusions.