research methods lec 2 Flashcards
define
correlation
occurs with or after
A in variable a occur with or after A in variable b.
define
causation
A variable a results in A variable b
change in variable A must occur is necessary for change in variable B
3 goals of scientific approach
description (qualitative)
prediction (correlational)
understanding-controlling (experimental)
psychology uses a ____ approach
scientific
scientific approach using ____ _____
empirical
experimental methologies
to understand phenomena of interest
define
non-experimental research
observational
cannot demonstrate causation
define
experimental research
can demonstrate causation due to its key features
manipulation
control of extraneous variables
define
IV
Independent variable= manipulated by experimenter
define
DV
Dependent variable
= participants response
in terms of variables, what do experiments ask
experiments investigate the effect of ____ on ___
whether systematic variation in one variable produces variation in another variable
IV on DV
define
observation
the quality/intensity/direction of the DV
define manipulation
the quality/intensity/direction of the IV change is controlled by the experimenter
change of _V => change in _V
this allows us to test the effect instead of waiting for it to
IV
DV
occur naturally
If Iv changed and DV is changed as a result then there is ______ about the ____ relationship
strong evidence
causal relationship cause DV occurs because/after IV
experimental group gets _V
control group get _V
experimental treatment = IV
control group = no treatment = no IV
RCT aka
randomised control trial
3rd variable aka
confounding variable
causes changes in the variable that is being measured
extraneous
unrelated
method to control extraneous variables (1)
1 = ensure all important factors are the same in both experimental and control group
validity _ types =
2 types internal validity (iv) external validity (ev)
define
internal validity
outcomes - consequences of the relationship between the IV and DV
no alternate explanations = therefore all extraneous variables are controlled for
therefore with confidence = change in IV = change in DV
define
external validity
- extent to which the results can be generalized
to the pop from the sample
OR
to the real world from experimental setting
IV => DV
relo outside experimental situation???
trade-off situation between ___ and ___
internal and external validity
^ iv =
high internal validity = more experimental control required = risk of compromising the ev
ev = results valid in the real world
effect of ___ on ___ occured by chance = explained by the ___ behind ____ ___
iv on ev
reasoning
inferential statistics
reasoning behind inferential statistics 3
3
1 assume there was no effect (Null hypothesis H.) groups aren’t diff
2 find the prob. of getting the result we got, given that the Null hypothesis H. is true
3 if the result is
pretty likely to have occurred by chance, then there is no reason to believe that there was an effect.
pretty unlikely to have occured by chance, then we are prepared to believe that there was an effect.
define
probability of an event
is the ratio of the no. of observations of that event over the no. of all observations of that event
define
Null Hypothesis H.
we don’t believe that there is an effect because many events may occur just by chance
hypothesis testing we assume that
estimate the probability of our result to have occurred by chance, assuming the Null hypothesis is true.
if probability of an event occuring =very low ___ %
less than 5.% = ready to believe that there is real effect and we reject the null Hypothesis
( we never accept Null hypothesis)
snd aka
signal detection theory
define hypothesis testing =
1
decisions are based on uncertain information. 4 possibilities = each expressed as a conditional probability
the probability of type 1 occurring is in fact controlled by the researcher
mainly interested in minimising the type 1 error
p-value indicates the probability of a type 1 error
there is no 100% certainty with statistics
alpha-level of hypothesis testing =
the alpha level represents the probabilty of a type 1 error = aka = level of significance
the accepted level in psychology =
medical sciences =
psychology = 5% alpha 0.05
medical sciences = 1% alpha 0.01
a level of significance of alpha 0.05 represents =
the researcher is willing to accept a 5% chance of making a type 1 error
when 2 pop.s are different the sample means =
tend to be larger for the larger pop ( pop with highest values)
define
t-distribution =
in small-medium samples the differences in sample means are described by the
t-dist symmetrical about zero
has bell-shaped curve
distribution of mean differences
assuming the two populations are the same ( H0)
therefore, can use the normal dist to find the probability of finding the differences between the sample means = if so = suggest that there is difference between the 2 pop.s
reasoning behind hypothesis testing =
extremes of normal dist = 2.5 each = total 5% of extreme values
region = region of rejection
if mean difference falls within the ___ region then __
since
region of rejection
we can reject H0
at alpha = 0.05
z > 1.966 or -1.96
t-test =
involves using a formula that returns a t-statistic = often called = t-obtained
t-obtained =
the t-statistic derived from the t-test
= difference in means / variability
non-experimental research pros=
good as pilot study = to inform where possible causal relationships could be
complex = real-world issues aren’t easily studied in lab
in many situations IV = can’t be manipulated
sometimes - unethical
need to design study groups that already exist via quasi-experimental designs
descriptive research =
seeks to describe phenomena rather than to manipulate variables
methods of descriptive research =
case studies
naturalistic observation
survey research
define
case studies
in-depth study of the behavior of 1 person in a small group
case studies = information=
used in clinical research (.e.g Freud)
used when large numbers of participants aren’t available
cons
small sample size
susceptibility to researcher bias
problem of a rare case
define
naturalistic observation
in-depth study of the phenomena in its natural setting
pros and cons of naturalistic observation=
pros
good generalisability
cons
observation per se can alter behavior
can’t establish cause of behaviour
define
survey research =
asks questions of large numbers of persons to gain information on attitudes and behaviours
2 approaches
approaches of survey research =
2
questionnaires
interviews
cons of survey research=
sampling issues
people may not respond accurately