# Confidence Intervals – CASP

< Back to search results-
**Format**Texts -
**Language/s**English -
**Target Audience**Researchers -
**EBM Stage**3 - Appraising evidence -
**Duration**<5 mins -
**Difficulty**Introductory

## Key Concepts addressed

## Details

The p-value itself gives no direct indication of how large or clinically important the estimated effect size is; it does not even tell you which treatment produced better outcomes in the trial. For this reason, confidence intervals are often preferred.

**Exercise**

How many possible outcomes are there from rolling a single dice? What is the probability of each of these outcomes? Plot each possible score against the probability that it occurs.

How many possible outcomes are there from rolling two dice? What is the probability of each of these outcomes? The easiest way to do this is to draw up a 6×6 table and fill in the total scores in each cell, and then count up how often they occur. Plot these probabilities on the same plot you created for rolling one dice, dividing the total score by two to get the average score per dice.

How many possible outcomes are there from rolling three dice? What is the probability of each of these outcomes? You can draw up 6 versions of the tables above and fill in the total scores including the third dice, or you might prefer to list all the combinations that add up to each of the possible scores from 3 to 18 and count how many there are. Plot these probabilities on the same plot you created for rolling one and two dice, dividing the total score by three to get the average score per dice.

## Discussion

0 Comments