Please enable JavaScript.
Coggle requires JavaScript to display documents.
quant W5-7 research skills p2 - Coggle Diagram
quant W5-7 research skills p2
distributions
bimodel distributions
bimodal- 2 peaks, multimodal- more than 2 peaks
negatively skewed distribution
the peal is shifted towards the right and tail extends to the left
outliers/extreme scores
impact of outliers
the mean is sensitive to extreme scores, outliers bias SD and error estimates
data points/scores that are very different to the rest of your data
types of research/ alternative hypothesis
non causal/associative
suggests particular charcteristics of behaviour without reference to causation
causal
suggests a particular causal infleunce
only appropriate if you are using an experimental design
kurtosis
shape of peak (flat/steep/shallow
kurtosis leptokurtic- very peaked
kurtosis playkurtic- low peak
kurtosis mesokurtic- middle peak
skewness
the extent to which your frequency histogram is lopsided rather than symmetrical
skew suggests data that deviates from symmetrical normal distribution
def- when we have a large set of data we must observe the distribution of the data
standard normal distribution
normally shaped distribution with a mean of 0 and a SD of 1
p value
happened by chance
if p greater than 0,05- accept the null hypothesis, reject alternate
null is significant when (calculations)-form hypothesis and collect data, run a statistical analyses on data to produce stats, compare stats with known distribution of values that allows us to work out how likely it is to obtain that statistic if there were no effect on population, if calculated p is less than 0.5 it suggests the pattern of findings is unlikely to be due to chance= statistically sig
if p is less than 0.05 accept alternate hypothesis and reject null
standard error
the SD of the sampling distribution of the mean
a measure of the degree to which the sample means deviate from the mean of the sample measn
must also tell us the degree to which the sample means deviate from the population mean
sampling
if you plot the sample means of many samples from one particular population
sampling error- sometimes its likely the patterns of scores in our samples dont accurately reflect the underlying population
z scores
tells us how mnay standard deviations above or below the mean score is
to calculate- subtract the sample mean from the score and then divide by the sample SD
ceiling effect
when a measure procedures most values near the top end of a scale (an easy test)
floored effect
when a measure produces most values near the bottom end of a sclae (a very hard test)
confidence interval (CI)
95% (typically)
we can use the sample mean as an estimate of the population mean
provides intervals that give us confidence in our sample mean and how it is close to the popualtion
