0. Statistical Analysis Flashcards
statistical analysis
part of mathematics that studies the collection, organization, analysis, interpretation, and representation of data
What to do if I want to find if the data is randomized
To find if data is normally distributed (randomly collected) I need to draw a distribution graph. If the graph is “bell-shaped” then the distribution is normal.
What percentage of data should fall within one standard deviation of the mean on a distribution graph?
Approximately 68%.
What percentage of data should fall within two standard deviations of the mean on a distribution graph?
Approximately 95%.
What percentage of data should fall within three standard deviations of the mean on a distribution graph?
Approximately 99.7%.
error bar
a vertical line below and above the mean value - it shows the variability of data
Error bars can show:
- The range of data
- a line extends from the highest to the lowest value of the data set - STDEV
- summarizes the spread of data around the mean value (plus minus 1 stand. deviation) - tells how widely the values are spread - the more reliable way of representing the variability of a data set than the range
two hypotheses in a t-test
- null/zero hypothesis (H0)
- alternative hypothesis (H1)
null hypothesis
always negative - claims that there is no significant difference between two sets, any observed difference is due to chance selection of data (claims the lack of effect of the investigated IV on the DV)
alternative hypothesis
claims that there is a significant difference between the two sets (the investigated IV does affect the DV)
“P” is the probability that…
…you’ll get the result you did if there is no significant difference between the 2 sets (the null hypothesis is correct)
- IV doesn’t impact the DP
If P<0.05…
…the sets are different (H1 is correct).
If P>0.05…
…the sets are the same (H0 is correct).
Even if the two bell shapped frequency distribution graphs overlap this can be determined as insignificant if…
…the values’ means are far apart. However, the P value must be calculated to confirm that they are not part of the same set.