Flashcards
Statistics
Flashcards do conjunto (28)
- 1Nominal measures- Categories are "nominated" with a number, e.g. 1=male and 2=female
- 2Ordinal measures- The numbers represent an "order", but the distance between 1st and 2nd is not necessarily the same as between 2nd and 3rd
- 3Interval measures- The numbers are actual and the distance between them constant, but there is no real 0, e.g. temperature
- 4Ratio measures (rati0)- The numbers are actual and the distance between them constant, and 0 means 0, e.g. correct answers
- 5Parametric statistical test- Used on interval and ratio measures, e.g. t-test or Pearson's correlation
- 6Non-parametric statistical test- Used on nominal and ordinal measures, e.g. Spearman's correlation
- 7Scatterplot- Used to show nominal measures on a figure
- 8Histogram/Frequency distribution graph- Shows the distribution of a phenomenon, used for interval and ratio measures
- 9Ideal-world histogram/Normal distribution- Bell-shaped, symmetrical, applies to many many phenomena
- Mean=median=mode - 10Skew- The distribution is asymmetrical and more dense on one side
- Normal distributions have skew=0
- Most graphs have -1>skew<1 - 11Outliers- Can be taken out if skew is more tha 1 or less than -1
- NOT in lepto- or platokurtic distributions
- If it doesn't help, make a non-parametric test instead - 12Positive skew Negative skew- Tail is moving away from 0
- Tail is moving towards 0 - 13Kurtosis Leptokurtic Mesokurtiv Platykurtic- Deformation of normal distribution
- High and thin (leap)
- Normal
- Low and flat (plat) - 14Central tendency Mean Median Mode- Summary of scores, shown with mean, median, and mode
- Gennemsnit, sum/number of scores
- Median, the middle score
- Most frequent score - 15Spread- How the data is distributed
- 16Range- Distance from smallest to largest score, used for ordinal, interval, and ratio measures, vulnerable to outliers
- 17Standard deviation- How much do scores differ from the mean on average? Used for interval and ratio measures
- √(Σ(X−M)^2)/(N−1) - 18Normal distribution and standard deviation in %34% of scores fall between M+1SD
34% of scores fall between M-1SD - 19Standard deviation and hypothesis testing- Mean A and mean B needs to be 2SD away to be significantly different
- 20Standard error- How means between groups of data differs from the mean of the means
- SD/√N
(larger N=smaller SE) - 21Correlation- A change in one variable is related to changes in another variable
- 22Pearson's R Spearman's Rho- Tests correlation, used for interval and ratio measures
- Tests correlation, used for ordinal measures - 23r>1 r=0 r<0- Positive correlation
- All data lies on the line
- Negative correlation - 24Spurious correlation- The group is not homogenous
- 25Cohen's effects- r>0,1 is a small effect
- r>0,3 is a medium effect
- r>0,5 is a large effect - 26P-value- P is the probability of something happening by chance
P=0 means 0%, that it can only happen by chance
P=1 means 100%, that it will def happen by chance - 27Type-1 error Type-2 error- Your data shows a correlation when there is none
- Your data does not show a correlation when there is one - 28Practical or social significance- When even a small significance is considered important, because it has a tremendous effect, e.g. in medicine
Informações
- Descrição
- Psychology FlashCards sobre Statistics, criado por Michella x em 04-05-2018.
- Criado por
- Michella x
- Visualizações
- 0
- Avaliação