Inferential Statistics Flashcards
What are inferential statistics?
They allow us to draw meaningful conclusions from our data, the data has to be quantitative and different types of quantitative data will be analysed by different statistical tests.
What do distribution curves show?
The spread of your data set
Normal distribution
The image on the right is an example of a normal distribution curve.
A normal distribution has the mean, mode and median in the centre, has 50% of the scores to the left and 50% to the right of the meanand is symmetrical.
normal = parametric tests
Standard deviation and the normal distribution
An important property of the normal distribution is the the proportion of scores falling either side of the mean is always the same.
Once the standard deviation has been calculated it will allow you to state the percentage of score falling either side of the mean. In a normal distribution:
- 68% of the scores lie within one SD from the mean
- 95% within 2 SD from the mean
- 99.7% within 3 SD from the mean
Skewed distributions
In a normal distribution, all the measures of central tendency lie together in the middle of the graph. This is not so in skewed distributions because there is a greater spread of scores on one side.
This therefore means that the rules about standard deviations do not apply to skewed distributions.
Skewed = non parametric
A useful way of looking at skewed distribution is to remember what each measure of tendency tells us.
- the MODE is the most frequent score, so is always the highest point of a distribution, whether it is normal or skewed.
- the MEDIAN is the middle point of the data, with 50% of the scores either side. In a skewed distribution the median will be towards the tail, as there will be more scores on the long side than the short side.
- the MEAN is the one most affected, because it takes the vlaue of every score into account, it lies furthest along the tail.
Level of significance and probability
You have looked at the use of summary tables & descriptive statistics in the previous section. From the use of means, modes, medians & grpahs we can make some conclusions about the differences between groups or variables but what we don’t know is how significant those differences are.
A psychologist cannot fully assess their findings until they have done some kind of statistical test on the data. Most psychological findings are assessed on the basis of probability.
Definition of probability in psychological research is..
The likelihood (probability) that a certain pattern in the data could be due to chnace rather than the actual variables being studied.
-By assessing the probability, we can determine the significance of the results.
Definition of significance in psychological research is…
- A significant result is one where there is low probability that chance factors were responsible for any difference or correlation between varibales tested.
- Therefore if significance is high the probability of the results being due to chance is low
- Whereas if the significance is low the probabilty of the results being due to chance is high.
- OVERALL as the probability of the results being due to chance increases, the significance of the results decreases.
Choosing a significance level
Psychologists can choose what level of probability they want to test their results at. In general psychologists agree that if there is less than 5% probabilitythat our results were due to chance, we can consider it significant
Level of significance written as ‘p’ values.
p<=0.05 meaning
the probability that results are due to chance is equal to or less than 5%
If psychologists wanted to be more certain about their results than this they could choose a stricter significance level, for example at..
P<=0.01 meaning…
That the probability that results are due to chance is equal to or less than 1%
P<=0.001 meaning…
That the probability that results are due to chance is equal to or less than 0.1%
