Skip to content

Information Value Calculation

Why It Matters

We over-invest in information gathering, either from curiosity or false diligence ("I need more data before deciding"). But information gathering has costs: time, money, delayed decision, analysis paralysis. Information value calculation provides discipline: only gather information if it might change your action and the expected benefit exceeds the cost. This prevents both over-analysis and premature decision-making.

The Key Move

Before seeking more information, ask: "What's my decision with current information?" Then: "What could I learn, and how would each piece of information change my decision?" Estimate:

  1. probability of getting each type of information,
  2. how it would shift your probabilities,
  3. whether it would change your decision,
  4. value difference if it does change your decision. If expected value of information (EVOI) exceeds cost, gather it. If not, decide now with current information.

Classic Application

Medical testing. Patient has moderate probability of disease (40%). Treatment has side effects, so you'll only treat if probability exceeds 60%. Should you order an expensive, invasive test? Calculate: If test is positive (30% chance given current beliefs), posterior jumps to 85% - you'd treat (different action). If negative (70% chance), posterior drops to 15% - you wouldn't treat (different action). Test would change your decision in both cases, so it has high information value. Order it. But if you'd treat regardless of test results (current 40% exceeds your treatment threshold), information value is zero - don't test.

Surprising Application

Career exploration before committing. You're deciding between job A and B, leaning 60-40 toward A. Should you spend three months doing informational interviews in field B? Calculate: What could you learn? How would it shift your probabilities? Would it change your decision? If you'd only choose B over A if probability exceeds 80%, and informational interviews would at best shift you to 55% (not enough to change your decision), they have low information value despite seeming "diligent." Either commit to A now, or set a clearer threshold that information could actually reach.

Failure Modes

Ignoring information costs - analysis paralysis occurs when you forget gathering information has costs. Underestimating decision urgency - sometimes delaying to gather information is the costliest choice. Overconfidence in current beliefs - if you're certain you're right, information value is zero; but that certainty might be miscalibrated. Pure option value - sometimes information creates new options, not just improves current choices; EVOI undervalues exploration.

Go Deeper

Raiffa & Schlaifer, Applied Statistical Decision Theory, Chapter 5 on information value; Howard, "Information Value Theory" (1966).

  • Option Value - Information Economics: Reasoning Under Asymmetry and Uncertainty