3. Experimental Statistics Flashcards
numerical data used to measure and describe characteristics of groups
descriptive statistics
to infer from sample data the probability of something being true of a population
inferential statistics
bar graph depicting frequency distribution
histogram
a graph cluster of two variables, w/ slope suggesting relationship (used for correlation studies)
scatterplots
(graph) goes up together and goes down together
postitive correlation
(graph) one goes up and one goes down
negative correlation
most frequently occurring score
mode
the average score
mean
the middle score in distribution (half above, half below)
median
difference b/w highest and lowest scores (can be skewed w/ outliers)
range
measure how much scores can vary around the mean (square root of sum of deviations divided by the number of scores)
standard deviation
what does each experiment have?
independent and dependent variables
experimenters manipulate what to observe what?
experimenters manipulate one or more factors (IV) to observe the effect on some behavior or mental process (DV)
in a study about flashcards: the flashcards would be the?
independent variable
in a study about flashcards: the test scores would be what?
dependent variables
what could impact research results as they are things that could mess up results; out of experimenters control
confounding variables
are variables that are anything that remain constant
control variables
what are important for experiments integrity?
random selection and random assignment
what firstly should be done to a diverse population you want to study?
a representative sample must be RANDOMLY SELECTED from that diverse population
what secondly should be done to a diverse population you want to study?
that randomly selected group should then also be be RANDOMLY ASSIGNED to the experimental and control groups
what does the random system ensure?
it ensures no bias or lack of diversity
what does a valid experiment need? (besides the random system)
it needs a large sample size
what’s wrong about small sample sizes?
they can produce inconsistent and abnormal results
the less confounding variables in an experiment has the better?
chance of the experiment reflecting the population.
