The interactive research resource for bioscience students
 Introduction > Step-by-Step Statistics > Gentle Introduction > Probabilities and Confidence Intervals |
Probabilities and Confidence Intervals The probability (often referred to at the P-value) gives an indication of the likelihood of your results occurring by chance. The P-value, measured on a scale of 0-1, will determine if you should accept or reject your null hypothesis. For example, if you role a dice, there is a 1 in 6 chance of you rolling a particular number. The probability of you rolling a 1, therefore is 0.167 (1/6). The probability of you rolling an even number is increased to 0.5 (3/6).A 'low' P-value (P<0.05) means that if you repeat your experiment 100 times, you should expect to record a similar result 95 times therefore the likelihood of your results occurring by chance is low. If you record P<0.05, you should reject your null hypothesis. If you record a high P-value (P>0.05), you should accept your null hypothesis. Be aware that although general convention is to accept P<0.05 as the 'cut-off' mark, some disciplines/journals accept P<0.1 as the probability convention. The P-value will be quoted by your statistical package. The confidence interval (CI) is often quoted as a range of numbers within which the majority of your data will fall. For example, Gillian is 95% confident that blackcurrant plants produce between 29 and 42 flowers per plant. Although 5% of plants will not conform to this, the majority of her plants will.
|
||
| <<< Previous Page >>><<< Next Page >>> |