COM5101 Chap.2 Testing and Comparing Communication Theories
Testing and Comparing Communication Theories
Criteria answer different questions
The course reading evaluates theory through explanatory power, predictive power, parsimony, falsifiability, internal consistency, heuristic value and organising power. These are not interchangeable virtues. A model can predict accurately while leaving the mechanism obscure. A detailed account can explain a small domain while forecasting poorly outside it.
Comparison begins by deciding which quality matters for the task.
Prediction and explanation
Prediction asks whether a theory anticipates an event or pattern. Explanation asks whether it offers a plausible account of why the pattern occurs. The distinction becomes decisive when intervention is proposed.
A predictor may direct attention to a likely outcome, but changing the predictor will not necessarily change the outcome. Intervention requires stronger confidence in the mechanism.
Parsimony with a condition
Simplicity is valuable only when competing accounts explain and predict equally well. Removing a real contextual moderator is not elegant; it is inaccurate.
Falsifiability adds a different discipline by requiring a proposition to face possible failure. Internal consistency asks whether the claims can be true together before evidence is even considered.
Productive theories can be incomplete
Heuristic value describes the ability to generate useful new hypotheses. Organising power describes the ability to bring existing knowledge into a coherent structure.
A theory may contribute strongly on these dimensions while remaining narrow or partly falsified. A mature evaluation therefore ranks strengths for a purpose instead of announcing a universal winner.
Compare on shared ground
Two isolated summaries do not amount to comparison. Put both theories against the same outcome and evidence.
State where their predictions diverge, then propose an observation that would make one account more plausible. Preserve the legitimate domain of the weaker account in the final verdict.
What this chapter covers
- 01
Explanatory power
- 02
Predictive power
- 03
Parsimony
- 04
Falsifiability
- 05
Internal consistency
- 06
Heuristic and organising value
Worked example · free
Compare arousal and identity accounts
- 1Name a shared outcome.
- 2State both mechanisms.
- 3Design divergent conditions.
- 2Give a bounded verdict.
Key terms
- Explanatory Power
- A theory's capacity to provide a plausible account of the phenomenon it was built to explain.
- Predictive Power
- A theory's capacity to anticipate events or patterns accurately under stated conditions.
- Parsimony
- Preference for the simpler account when competing theories explain and predict equally well.
- Falsifiability
- The capacity of a proposition to encounter observations that could count against it.
Testing and Comparing Communication Theories FAQ
Can the most accurate theory still be a weak explanation?
Yes. A statistical pattern may forecast sharing or attention without identifying the process that generates it. That can be useful for prediction but inadequate for intervention. If the task is to change an outcome, explain why the observed predictor is connected to the mechanism rather than assuming accuracy supplies causation.
Does parsimony mean the shortest theory wins?
No. Simplicity is preferred when competing accounts explain and predict equally well. A theory that removes a real boundary or moderator becomes simpler by losing information. Evaluate whether the extra complexity changes a decision, prediction or mechanism before treating it as unnecessary.
How do I finish a theory comparison?
Give a bounded verdict linked to the criterion and task. State which account is stronger for the present decision, why the evidence favours it, and where the rival remains useful. Avoid the empty conclusion that both theories have strengths because it leaves the comparison unresolved.
Can a partly falsified theory remain useful?
Yes, if its surviving propositions organise evidence or generate productive tests. State which claim failed, what revision followed and which domain remains defensible. Do not treat heuristic value as proof of accuracy, but do not confuse one failed proposition with the disappearance of every useful relationship.
Assessment move
Build a criteria table, but do not assign invented numerical scores. For each theory, write one sentence on explanation, prediction, testability and boundary. Then identify the criterion that controls the question you are answering. Practise comparisons with a shared case so that the accounts cannot pass each other without contact.
For closed book recall, memorise each criterion through a diagnostic question: What does it explain? What does it predict? What could disprove it? What knowledge does it organise? What new question does it generate? End every practice response with a decision tied to one criterion. Use a two-pass comparison drill. On the first pass, argue for the account with stronger prediction and state what it leaves unexplained.
On the second, argue for the account with the clearer mechanism and state what it predicts less well. Compare the verdicts and identify which practical decision makes one criterion more important. Keep a short falsification ledger beside the table. Record one result each theory would struggle to absorb and the revision that result would force.
This prevents flexibility from being mistaken for strength and gives you a precise way to discuss theories that remain productive after some claims fail.
