week 10: variance + inferential statistics Flashcards
what does an experiment ask?
does variation in the IV cause variation in the DV
how do we combat sources of variability?
isolating the effects of the IV
example of isolating effects of independent variable
when measuring the effect of diversity training on implicit racial bias, we do not care about effect of being younger vs older or scale of caring.
what are sources of variability in this example?
different values of diversity
differences in age
differences in mood
differences in SES
chance factors
what is systematic variance?
variability caused by manipulating the IV (anti-bias diversity training)
what is error variance?
variability caused by everything else
what are examples of error variance?
- individual differences
-random factors
-misunderstood instructions
-unexpected events
-experimenter/equipment errors
what makes up inferential statistics?
systematic variance + error variance = total variance
what is total variance?
total variability in the DV caused by all factors
what are ways of dealing with potential confounds?
avoid them !!
turn it into a control variable
what is a control variable?
a factor that is fixed throughout the experiment
if it cant vary then it cant introduce variability in the DV
what is a control variable that be use in the diversity experiment?
age group, comparing 40+ and less than 40
what is a drawback of too much control?
limited generalizability if the findings
make the potential confound a…
random variable
what is a random variable?
a factor that varies randomly and unsystematically throughout
if we can ensure that it varies randomly, it will contribute to….
error variance
inferential statistics asks
are differences “significantly different”
what determines statistical significance?
- magnitude of difference, that is due to systematic difference
- variability within groups
statistical significance increases with ……..
and decreases with ……….
- mean difference
- variability
what are methods to measure statistical significance and what does it measure?
t test, measures if there is a significant difference between the means of 2 groups and how they are related
I t I > 1.96 ->
statistically significant
determining statistical significance is a probabilistic judgment
what is the significance p= .05
alpha criterion, normal threshold in psychology.
if p value is less than .05 than mean difference is statistically significant
what is type I error?
false positive
replication is important
what is type II error
false negative
IV with two “levels”
- Anti-bias diversity training condition
2.No training condition (control condition)
what is the yerkes-dodson curve
it proposes that you reach your peak level of performance with an intermediate level of stress, or arousal
what is a ceiling effect?
extreme high values
what is floor effect
extremely low values
how to avoid ceiling and floor effects?
- pilot studies
- multiple IV levels
- noticeable difference in IV levels