U2T8 - Keywords Flashcards
Kingdom Animalia
Null Hypothesis
There is no significant difference between between A + B. Any observed difference is due to chance.
Sample Mean
Mathematical average. In normal distribution, mean, median + mode all equal. May not truly be representative of entire pop. x̅
Modal Value
Most common value.
Median Value
Central value in data set.
Standard Deviation
Variation of data set. Smaller value means closer to mean. In normally distributed data set, value covers 68% of samples either side of mean. σ
Standard Error
See how close sample mean is to pop mean. AKA standard deviation of the mean. If small, diff is small so sample mean reliable. σx̅
95% Confidence Limits
Limits between which true mean of pop almost certainly lies. Find t by reading p = 0.05 at n - 1 degrees of freedom.
Error Bars
Show 95% confidence limits on a graph. If bars overlap, no significant difference.
P
Probability of outcome being due to chance. Lower p value means less likely to be due to chance.
Critical Value of Probability
p = 0.05. i.e. 5% Means due to chance. Less and it’s not likely to be due to chance.
Student’s t Test
Allows comparison of 2 sample means. Allows us to asses whether diff between 2 sample means is significant or due to chance. Must use t probability table to determine whether null hypothesis may be accepted or rejected. If prob value more than 0.05, null hypothesis accepted.
p>0.05
No significant diff. Due to chance. Null accepted.
0.05>p>0.01
Significantly diff at 95% probability but not 99%. Significant diff. Null rejected.
0.01>p>0.001
Significantly diff at 99% prob but not 99.9%. Highly signif diff. Null rejected.
p<0.001
Significantly diff at 99.9% prob. Very highly signif diff. Null rejected.