Week 2:Intro To Stats Flashcards
What is Quantitative Research?
-collects numerical data analysed by stats to explain phenomena
-It can use quantities, surveys, tasks e.g. reaction time tasks,experiments etc.
What is Qualitative Research?
-it’s exploratory to help us gain understanding of underlying reason
-opinions,motives,asking people,interviews,focus groups etc.
Define a sample
a group of participants from the target population (matches general characteristics to make it representative) where the findings can be generalised.
What can be the issue with samples?
it’s not easy because we can’t be 100% confident that the findings will fit the whole population so we can be limited by the sample we use.
Define Within-subjects design
essentially repeated measures design (doing both tasks)
What are the advantages of within-subject design?
-good to use when number of resources/participants are limited
-good when studying a real life setting
Define Between-subjects design
essentially independent group design (only doing one condition)
What are the pros/cons of between-subjects design?
P-shorter study duration prevents carry over effect of learning/fatigue
C-larger number of participants required for higher statistical power
Define the IV/DV
IV-what the researcher manipulates
DV-what the researcher measures and can change based off the IV
Define nominal data
categorical data e.g. male/female, blue eyes/brown eyes etc. (qualitative data NOT quantitative)
Define ordinal data
ordered data presented in rank order but the intervals between data is not necessarily the same e.g. shoe size,level of attractiveness etc.
Define interval data
data has equal intervals between points measured in fixed units and can also go into the minuses but has no ‘true zero’ e.g. temperature,voltage etc.
Define ratio data
similar to interval data with equal intervals measured at a continuous scale but has a ‘true zero’ e.g. height,weight,scores on a test etc.
Explain what descriptive statistics are
-describes the data letting us see what it looks like
Two types-measure of central tendency AND measures of dispersion/spread(you report these together)
-only tells us about the sample not really the population
Explain inferential statistics
random sample of data taken from the population so we can make inferences about the population and reach conclusions that reach beyond our data
How do we determine what statistic to use?
we base it off the distribution and data level (NOIR) of our data
Define central tendency
describes how ‘most’ people behave and can involve the mean,median and mode (5% trimmed mean)
Why can the mode be problematic?
It’s mainly used for nominal data as its the most common value BUT it ignores a lot of information about the other scores
Define the median
central value usually used by ordinal/skewed data and only uses one/two values which can sometimes be problematic
Why can the mean be problematic?
it can be affected by extreme values due to an outlier
Explain what a normal distribution looks like (histogram)
-majority of values in the middle with ‘bell-shaped curve’
-few extreme scores on either side
Explain what a non-normal distribution looks like(histogram)
-majority of scores at one end of the distribution
-skewed data so no longer central meaning extreme scores affect the mean
How do we get around the problem of extreme scores?
-trim our data could use 5% trimmed mean (take 5% of scores from each end high/low) tend to use the median as a method more
-makes it more representative
Give examples of measures of dispersion/spread
-the range
-the interquartile range(the difference between the middle 50% of scores)
-variance and standard deviation(the average distance of scores from the mean)
How can the range be problematic?
-it can be affected by extreme scores because it only takes into account the highest/low score
-BUT if our data is skewed we can use the interquartile range
How can we calculate variance?
subtract the mean from each score so we need to find the average of these
Why can’t we make inferences with descriptive statistics?
■ Although one group may differ from the other (have a greater measure of central tendency etc.), this doesn’t mean that there is actually a difference as it could just happen by chance
■ Descriptive statistics don’t tell us whether the difference between these groups can be inferred beyond our sample to the population
■ What we need is inferential statistics and we need to look at probability (p) values
