General statistic informations Flashcards
Reliability vs. validity
- helps measure credibility
Reliability: consistency of the experiment results, if an experiment is reliable it should produce similar results every time. ( also known as reliability)
Validity: refer to the accuracy of a study or measure, reliability is usually necessary but not sufficient for validity. producing the same results as reliability does not mean that your result reflects reality thus fora study to be well considered you want it to have both reliability and validity.
Explain the concept of Error bars
- They represent several types of statistics: confidence intervals, standard deviations, standard errors.
What is confidence intervals: This tells us how good a sample in a study is representative of the actual population. It does that by providing a range of values that emulates what we are bound to see in the population
- CI can help to determine statistical significance: whenever CI do not overlap the group mean is statistically significantly different. the opposite is not always true.
What is regression?
- All variables examined are continous
What is correlation?
- 1 correlation is good
- -1 correlation is opposite
- 0 no correlation
chi-squares?
- when you can categories your variables, you can look at them in 2 distribution and see how they differ from each other
- null vs. alternative hypothesis
What is a T-test?
- compare the mean of a group to a specific value