Psychology DATA TEST

ATAR, DATA TEST, Inferential statistics

The Scientific Method

The Scientific Method

a logical process of problem-solving applied in all sciences

IV (independent Variable)

the variable that is manipulated by the experimenter resulting in changes in the dependent variable (DV)

DV (dependent variable)

the property that is measured in psychological research, to look for effects of the independent variable (IV)

operationalization

quantification of a variable (put into numbers)

extraneous variable

a variable other than the IV that could cause changes in the value of the dependent variable

confounding variable

a variable other than the IV that has a systematic effect on the valyd of the dependent variable; if a confounding variable exists, no valid conclusions are drawn from the data

hypothesis

a prediction of the outcome of research stated in terms of the influence of changes int he value of the IV on the value of the DV

null hypothesis

there will be no significant difference between groups; any difference observed is caused by error

alternate hypothesis

there will be a R’ship between the IV and DV

population

the group of people about which we wish to draw conclusions

sample

the members of the pop. who have been chosen to take part in the research

convenience sampling

using whoever is available at the time of research

random sampling

a sampling procedure in which every member of the pop. has an equal change of being selected

stratified sampling

sampling process by which the effects of a certain variable can be eliminated as a possible confound in an experiment

E-group

the group of research participants exposed to the IV; the results are compared with the C group so that the effects of the IV can be determined

C-group

the group of research participants not exposed to variations in the IV; the results are compared with the E-group to see/determine the effects of the IV

random allocation

subject-selection procedure where all participants who have been selected for an experiment have an equal chance of being in the E or C group

repeated measures design

subject-selection procedure where each participant is part of both the E and C group

counterbalancing

a method for controlling order effects in a repeated measures design

matched participants design

subject-selection procedure that attempts to eliminate confounding variables by ‘matching’ on key characteristics, each individual in the E group with an individual in the C group

placebo effect

participants behavior that is influenced by their expectations of how they should behave, caused by the belief that they have received some treatment.

single blind procedure

allocating participants to groups in such a way that they do not know whether they are in the E or C group

Experimenter effect

outcome of an experiment being influenced by the person conducting the experiment

double blind procedure

method of allocating participants to groups so that the experimenter and participants don’t know which group is which

naturalistic observation

observation of voluntary behaviors within a structured environment such as a lab

controlled observation

observation of voluntary behaviors within a structured environment such as a lab

interview

structured (asked a set of pre-determined questions with fixed choice of responses) or clinical (structured guidelines, but further questioning is used for clarification)

surveys (form of questioner)

question-and -answer response (rating scales). easy to replicate and score, provide a way of quantifying data. weakness - open to bias if participants are trying to appear in a particular way

psychological test (type of questionare)

multiple choice IQ tests and personality tests. strengths - standardized, easy to replicate, and score. weakness - difficult to contrast and validate

Generlisation

for it to occur: results must show statistical significance, all sampling procedures were appropriate, all experimental procedures were appropriate, all measures were valid, all possible confounding variables were controlled.

line graph/table

variable is continuous (can have any value within a certain range)

normal curve

bell curve. statistical procedures can be applied to the bell curve without further manipulation of data

measures of central tendency

mean, median mode - for a normal bell curve. how the data is clustered near the central point of the data set

measures of variability (dispersion)

range, variance, SD

range

difference between highest and lowest score on data set

variance

how much, on average, the scores differ from the mean

standard deviation (SD)

measure that tells you how far on average the scores differ from the mean

p value

probability value that the scores were caused by chance

p value of less than 0.05

statistically significant. less chance that the data was caused due to chance

p value of more than 0.05

not statically significant. caused by chance most likely

students t-test

a statistical test used to test whether the difference between the response of two groups is statistically significant or not. It is appropriate when the data sets follow a normal distribution, the population variance is unknown, and the sample size is relatively small (usually n ≤ 30).

Mann-Whitney U test

test of null hypothesis. does not require normal distributions. compare the differences between 2 groups when the DV is either ordinal or continuous

Wilcoxen Signed Rank Test

used as an alternative for t-test when the pop cannot be assumed to be normally distributed. compares 2 sets of scores from the same population

persons product moment correlation

measure of the strength of the linear r’ship between to continuous variables

spearman correlation

measure of 2 ordinal variables that use the ranked values for each variable to examine how they change together, but not necessarily at a constant rate

correlation

statistical measure of the strength of the r’ship between two variables, does not show a cause-and-effect r’ship, but describes the way in which variables vary in relation to each other

reliability

the extent to which a measure could be expected to produce the same result with the same subject(s) under the same conditions on other occasions

validity

the extent to which an instrument measures what it is supposed to measure

internal reliability

extent to which all items in a research instrument contribute = to the final score. High if the correlation between scores on the odd-numbered items and even-numbered items were high

inter-rater reliability

the same result should be obtained by anyone administering the test. procedures must be standardized

parallel form reliability

some tests have more than one form that measures the same property. these can be very useful if the research is investigating change in the property measured. measured before (pretest) and with a parallel form after (post-test)

test-retest reliability

checked to ensure that it would produce the same result if re-administered to the same person under the same conditions at a different time

internal validity

examines whether the results gained from a measure are truly due to the variable that it is thought to be measuring

content validity

face validity. form of internal validity that involves examining the instrument to decide if it measures what its supposed to

construct validity

form of internal validity. deciding whether the test can be used to support the theory that is being tested

external validity

criterion-related validity that refers to the extent that results from this measure are comparable with other, established measures of the variable

ordinal scale

data that has definite sequence but the gap between one level and the next is not consistent e.g. age of people in the classroom

nominal scale

data that has a QUALITATIVE value rather than quantitative value where there is no ranking or ordering of values implied e.g. hair colour

Interval scale

Data is measured on a scale, where each step is the same value but 0 does not mean the property does not exist (e.g. Celsius and Fahrenheit).

Ratio scale

measurements that represent quantities is terms of = intervals and an absolute 0 point of origin