Final Exam Flashcards
(231 cards)
1
Q
What are the essential aspects of an experiment?
A
- IV
-DV - Control (keeps extraneous variables consistent)
- Random Assignment
- Control Group
- Experimental group
2
Q
What is achieved when your experiment has good control?
what is control?
A
control = when all other variables other than the IV and DV are held consistent between the two groups
achieve = you can say that the IV caused the DV, and not extraneous variables
3
Q
what is it called when every participant has an equal likelihood of being assigned to either the experimental or control group?
A
random assignment
4
Q
what is the goal of random assignment?
A
to neutralize individual differences, making the two groups essentially the same
5
Q
What is a variable that varies along with the IV (due to a lack of control) called?
This variable can serve as an alternative explanation to changes in the DV
A
confound variable
6
Q
do confound variables discredit the experiment?
why?
A
YES
because they serve as an alternative explanation as to why there is a change in the DV that may not be the IV
7
Q
What is internal validity?
A
the extent to which the experimenter controls for confound variables.
means that your experiment measured and tested what it was supposed to. (can say that IV caused DV)
ability to say that you tested your hypothesis
8
Q
are confounds a threat to internal validity?
A
duh
9
Q
what are the 4 threats to internal validity?
A
1) Selection
2) Instrumentation
3) Experimental mortality
4) Experimental Bias
10
Q
Threats to internal validity:
what is selection?
A
if the two groups are different somehow before the experiment begins.
11
Q
Threats to internal validity:
what is instrumentation
A
changes in criteria used by observers / changes in mechanical measuring device
12
Q
Threats to internal validity:
what is experimental morality?
A
loss of subjects in an experiment.
if loss is different ACROSS GROUPS then the study will lack internal validity.
13
Q
Threats to internal validity:
what is experimental bias?
A
expectations of an outcome by persons running an experiment may significantly influence the outcome
14
Q
what is generalizability?
A
ones ability to say something about a population based on an observed sample…
kinda same thing as external validity
15
Q
what is external validity?
A
the extent to which the results of an observation generalize to other situations or are representative of real life.
16
Q
the degree to which the results of study can be extended beyond the research setting is called what?
A
external validity
17
Q
what are 4 threats to external validity?
A
1) lack of random sampling
2) mortality
3) artificial lab settings
4) reactivity on part of the subjects
18
Q
threats to external validity:
define: mortality
why is it a threat
A
loss of participants from a study
if the subjects who drop off are significantly different than thoe who remain the sample may not be representative of the population
(this is called SELECTIVE ATTRITION)
19
Q
threats to external validity:
why is an artificial lab setting bad?
A
data obtained in tightly controlled lab settings may not generalize to natural settings
20
Q
what is the tradeoff between internal and external validity?
A
high internal validity from tight control lowers external validity…
external validity means it applies the findings in the experiment extend to the real world..
but if the experiment is tightly controlled (has high internal validity) then its not really representative of real life.
21
Q
Word of the day: Confound
A
varies with the IV and can serve as an alternative explanation for the change in the DV
22
Q
if she asks “is this a true experiment” what is the one thing that you look for?
A
random assignment
23
Q
define: power
A
the ability to detect the effects of the Iv if they are actually there
24
Q
when do we use a Chi Square test?
A
when we want to test the independence of two variables
is x independent of y?
25
Chi Square:
what is the null hypothesis?
Variables x and y are independent.
26
Chi square:
what is the alternative hypothesis?
Variables X and Y are NOT independent
27
is Chi square a kind of statistical test?
yes!
its an inferential stats test
28
what is error variance?
- the variability in scores caused by variables other than your IV
(e.g. some subjects might be tired)
- can be extraneous or subject related variables
29
How do you handle error variance?
3 ways + what each thing means
1) reduce error variance
--> hold extraneous variables consistent
--> match subjects on characteristics
--> use within- subject design
2) Increase the effectiveness of your IV
--> use a strong manipulation
--> use sensitive DV
3) Randomize Error Variance
--> use random assignment
30
what is between-subject designs?
- participants are randomly assigned to the groups
- groups are independent
- different participants in both groups
- its also called independent group design
- its the type of design we do.
31
what is a within-subject design?
- one participant goes through all levels of the IV
- also called repeated measures design
32
what are the advantages of within subject designs?
- subject-related factors are literally identical across conditions, so there are not large differences to obscure the effects of the IV
- reduced error variance leads to a more powerful design
- serving as a control for yourself, more likely to find the effect of an IV because there is not as much error variance across groups
33
what are the disadvantages of a within subject design?
- more demanding on subjects
- there is a risk of carryover effects (order effects) where the previous treatment alters the behaviour in a subsequent treatment
34
What is the advantages and disadvantages of a between-group design?
-- requires more participants to have high statistical significance?
35
What are the two types of Between-subject designs?
each type has groups within it...
break em down.
1) single-factor randomized group designs
a) randomized two group design
b) randomized multi group design
2) matched group design
a) matched pair design
b) matched multi group design.
36
ok ok..
what is a randomized two-group design?
what are the advantages and disadvantages?
- randomly assign participants to 2 groups.
advantage: easy to conduct, don't need a lot of subjects
disadvantage: provides limited information
37
what is randomized multi group design?
three or more levels of the IV
subjects are randomly assigned
38
what are the advantages / disadvantages to randomized multi group design?
disadvantage: greater number of subjects needed
advantage: obtain more information
39
word of the day: Factor
means the same thing as the IV - they are interchangeable
40
what is a matched pair design?
only two groups are tested
- matched on a certain variable that is important to the experiment
e.g. equalize participants based on math ability if you are doing a algebra experiment
41
what is matched multi group design?
multiple groups are tested
42
what are the advantages / disadvantages for matched designs?
advantage: can control subject variables that may obscure the effect of the IV
disadvantage: more demanding and time consuming
43
carry over effects can happen in 'Within subject designs'
what are carryover effects (order effects)?
where a previous treatment alters the behaviour in a subsequent treatment
(how you behave in the first condition could effect how you behave in the next)
44
what are the sources of carry over effects?
x6
Leonard farted happily sensing completion arrival
1) Learning
2) Fatigue
3) Habituation
4) Sensitization
5) Contrast
6) Adaptation
45
carry over effect:
What is learning?
learning to perform a task in the first treatment may affect later treatments
46
carry over effect:
What is fatigue?
may cause performance in later treatments to deteriorate
47
carry over effect:
What is habituation?
repeated exposure to a stimulus can lead to reduced responsiveness
48
carry over effect:
What is sensitization?
repeated exposure to a stimulus can lead to increased responsiveness
49
carry over effect:
What is contrast
exposure to one condition may alter responses in later conditions
50
carry over effect:
What is adaptation
subjects may adapt to certain manipulations (e.g. drug trials)
51
how do we fix the issue of carry over effects?
counterbalancing
52
what is counterbalancing?
assign treatments in a different order for different subjects
53
what are the two different kinds of counter balancing?
complete counterbalancing and partial counterbalancing
54
what is complete counterbalancing?
provides every possible ordering of treatments and assigns at least one subject to each ordering....
e.g. 3 conditions --> 6 possible orderings
ABC CBA ACB BCA CAB BAC
55
what is partial counterbalancing?
what is the benefit?
includes only some of the possible treatment orders..
- each treatment appears equally often in each position.
- the benefit is not as many treatment orders need to be tested
e.g. treatment A appears 3 times in position 1, 2 & 3
treatment B appears 3 times in positions 1, 2, and 3
ect.
56
what is a within subject design?
each subject is exposed to ALL levels of the IV
also called repeated measures because participants are tested repeatedly across all conditions
(serves as a control fro yourself) more likely to find the effect of an IV because there is not as much error across groups
57
what are the advantages of a within subject design?
- all subject related factors are literally identical across conditions so there cannot be large individual differences to obscure the effect of the IV
- reduces error variance and leads to a more powerful design (more sensitive to the effects of the IV)
58
what are the disadvantages of a within subject design?
- more demanding on subjects
- there are risks of carryover effects
59
when would you use a within-subject design?
- when subject differences contribute heavily to variation in the DV
- when the number of subjects is limited and carryover effects are not an issue
(e.g. patients with bipolar [limited participants])
60
what are the two types of within-subject designs?
1) single factor two level design
2) single-factor multilevel design
61
within subject design:
what is a single factor two level design?
includes just 2 levels of the IV
all subjects receive both levels of the IV, but half receive the treatments in order one and half in the opposite order
62
within subject design:
what is a single-factor multilevel design?
a single group of students is exposed to three or more levels of a single IV
63
discuss the differences between a posttest-only design and a pretest-posttest design...
A posttest is an assessment measure given to participants after they have received treatment as part of a research study.
A pretest-posttest research design must provide participants with the same assessment measures before and after treatment in order to determine if any changes can be connected to the treatment.
64
what would you expect to find in the abstract section of a research paper?
- brief summary of paper
- the hypothesis
- the experimental method (materials, data gathering procedures)
- the findings as they relate to the hypothesis
- the conclusions / implications of findings
65
what would you expect to find in the introduction of the research paper?
- describes the problem under investigation
- literature review
- discusses methodology of study
- state the prediction of what they think will happen
66
what would you expect to find in the methods section?
what are the 4 sections in the methods section
- describes in detail how the study was conducted
- allows the reader to evaluate the procedures and replicate if they wanted to..
4 sections:
1) participants
2) Materials
3) procedure
4) design
67
what are the things that you would expect to find in the results section of a research paper?
- summary of the data collected
- the presented facts (they are not interpreted)
- descriptive stats
- a statement to whether the findings are statistically significant and supported the hypothesis
68
what would you expect to find in the discussion section of the research paper
- evaluates and interprets the results in terms of the original hypothesis
- discusses any difficulties of research
- discusses future directions / practical implications of findings
69
How is the References section layer out?
- alphabetical order
- indented
- double spaced
70
what is the appendix section used for?
- detailed description of materials
- helps the reader to understand, evaluate, and replicate the study
71
how is an in-text citation layed out?
(Jones, 1998)
(Jones & Charles, 1998)
(Jones et al., 1998)
72
how to cite a journal article in APA
- Author or authors. The surname is followed by first initials.
- Year of publication of the article (in round brackets).
- Article title.
- Journal title (in italics).
- Volume of journal (in italics).
- Issue number of journal in round brackets (no italics).
- Page range of article.
- DOI or URL
e.g.
Ruxton, C. (2016). Tea: Hydration and other health benefits. Primary Health Care, 26(8), 34-42. https://doi.org/10.7748/phc.2016.e1162
73
how to cite a book in APA
- Author or authors. The surname is followed by first initials.
- Year of publication of the book (in round brackets).
- Book title (in italics).
- Edition (in round brackets), if other than first.
- Publisher.
e.g.
Fletcher, D. P. (2018). Disrupters: Success strategies for women who break the mold. Entrepreneur Press.
74
What are factorial designs?
Two or more factors, each with two or more levels, are manipulated in a cross manner
they have more than one IV and each IV is present at every level of the other IV
75
What is the term factor synonymous with?
independent variable
76
What is the simplest factorial design?
a 2X2 between-subject factorial design
77
What is a 2X2 between-subject factorial design?
two factors with two levels of each IV, manipulated between subjects
78
What is the benefit of using a factorial design over two separate single-factor designs?
examine the effects of more than one IV, both individually and collectively, on the DV.
79
How many hypotheses does a 2X2 factorial design test?
3 hypotheses
80
What are the three hypotheses that a 2X2 factorial design tests?
1. Is there a main effect of factor A?
2. Is there a main effect of factor B?
3. Is there an interaction between factors A and B?
81
What is the main effect of factorial designs?
the separate effect of each IV
82
What is the separate effect of each IV (main effects)?
It is the effect of each variable separately on the DV (like doing separate experiments with single IVs)
83
How to determine the main effects?
Compare row means collapsed across columns (main effect of Factor A)
Compare column means collapsed across rows (main effect of Factor B)
84
What is the interaction of factorial designs?
when the effect of one IV is different at different levels of the other IV
85
Can interactions be determined by performing separate single-factor studies?
No, they are unique to factorial designs
86
What are the three possible outcomes when analyzing a factorial experiment?
1. There may or may not be a significant main effect for variable A
2. There may or may not be a main effect for variable B
3. There may or may not be a significant interaction between the two IV's
87
What do the lines on a graph look like when an interaction may be present in a factorial design?
not parallel
88
What do the lines on a graph look like in a factorial design when the effect of one factor on the DV is different across the different levels of the other factor? (aka an interaction)
non-parallel lines
89
what is a major distinction between qualitative and quantitative methods?
Quantitative techniques use __________ while qualitative techniques involve __________
numerical descriptions; verbal descriptions
90
What do the lines on a graph look like in a factorial design when one factor behaves the same at each level of the other factor (indicating NO interaction)?
parallel
91
close ended questions
a) give a fixed number of responsive alternatives
b) are difficult to code
c) give more information than open ended questions
d) are a good way to find out what people think
a) give a fixed number of response alternatives
92
What is a factorial within-subjects design?
a design when each subject is exposed to every combination of levels of all the factors
93
in order to eliminate the problem of a person always agreeing or disagreeing on all items in a questionnaire, a researcher would
a) use a more reliable questionnaire
b) instruct participants not to agree or disagree consistently
c) word half of the questions in a positive way and half of the questions in a negative way
d) use a more valid questionaire
c) word half of the questions in a positive way and half of the questions in a negative way
94
what are the disadvantages of factorial within-subject designs?
They require a great deal from subjects.
You need to control for carryover effects (counterbalance).
With more complicated designs, complete counterbalancing may become impractical.
95
for data measured on a nominal scale, you would use ______ for your statistical test
a) sd
b) t test
c) r
d) X2
X2
96
What is a higher-order factorial design?
the factorial design can include any number of levels of a given factor, and any number of factors
97
What do you need to consider when using a higher-order factorial design?
you need to consider the number of subjects required and the complexity of potential interactions
98
the correlation method involves
a) measurement of the participants on two variables
b) manipulation of an independant variable
c) elimination of the third-variable problem
a) measurement of participants on two variables
99
"do you think its important to decrease the extreme amount of money wasted on holiday decorations" is an example of a _________ question
a) double-barreled
b) double negative
c) simplistic
d) loaded
d) loaded
100
a confound variable is devastating to an experimental design because it
a) increases the variability in dada
b) increases the internal reactivity of the experiment
c) makes possible alternative explanations for the results
d) eliminates alternative explanations for the results
c) makes possible alternative explanations for the results
101
suppose you were to test the hypothesis that viewing pornographic material causes violence behaviour. the independent variable would be ______ and the dependant variable would be ___________
a) violent behaviour ; viewing pornographic material
b) viewing pornographic material ; violent behaviour
B
102
correlation studies allow researchers to ______ variables, experiments allow researchers to _______ variables
a) make predictions about ; explain relationships
b) manipulate ; measure
A
103
manipulating the independant variable involves
a) exposing subjects to at least 1 level of the independent variable
b) measuring two or more behaviours and determining whether they covary
c) exposing subjects to at least 2 levels of the independent variable
C
104
What are the main effects and interactions we obtain from a 2x2x2 factorial design?
A, B, C, main effects
AxB, BxC, and AxC 2-way interactions
AxBxC 3-way interaction
105
What is an ANOVA? What do we measure with this?
Analysis of Variance
Data from factorial designs
106
What does the ANOVA ask?
Is there a main effect of factor A?
is there a main effect of factor B?
Is there an AxB interaction?
107
What is a pre-test post-test design?
When you use a pretest to measure participants on the DV, then introduce the treatment, and then use the posttest to measure participants on the DV again to see if participants improved from the treatment
pretest -> treatment -> posttest
108
What are pre-test-post-test designs used for?
to evaluate the effects of some manipulation on subsequent performance
109
What do you need in a pretest-posttest design?
need a control condition
** the pretest and posttest are given at the same time intervals
110
What is the problem with pretest-posttest designs?
giving participants the pretest may change the way they perform after you introduce manipulation.
***Internal Validity COMPROMISED
111
What is the problem with pretest?
having a pretest may lead participants to perform differently than they would have without the pretest (EXTERNAL VALIDITY COMPROMISED)
112
What are developmental designs?
used to evaluate changes in behaviour that relate to changes in chronological age
113
what are longitudinal designs?
measurements are obtained from the same subjects over a period of time, often many years
114
What is an advantage of longitudinal designs?
they provide a great deal of information about intraindividual change (changes within a person)
115
What are disadvantages of longitudinal designs?
require a great deal of time and effort
subject mortality
carryover effects
116
What is subject mortality?
- the loss of participants from the study; if those subjects who drop out are significantly different from those who remain (called selective attrition) the sample may be biased
117
What are carryover effects?
improved performance on tests over time may be related to the participants' increasing experience with taking the tests than to changes related to age per se
118
what are cross-sectional designs?
all measurements are performed at about the same time, but are done on individuals of different ages
119
What is a disadvantage of cross-sectional designs?
they confound the effect of age and culture
120
What is an advantage of cross-sectional designs?
easier and less time consuming and less costly than longitudinal studies
121
What are cohort or generation effects?
when behaviour (your DV) is influenced in some way by the generation in which one was born, and by the corresponding social and historical forces
122
what is a cohort?
a group of people who share a commonality, usually age
123
What is the cohort effect?
when age is confounded with the period and historical forces that a group experienced (we don’t know whether it is age per se or the time period in which the group grew up that affects the DV)
124
what is a sequential design?
blends both the longitudinal and cross-sectional designs.
125
What is an advantage of sequential designs?
it does not take as long as longitudinal studies and the groups are typically closer in age so the study is less likely suffer from cohort effects
126
What is a quasi-experimental design?
These designs resemble experiments but lack important features of true experiments.
Because they lack random assignment, manipulation of an IV, and control they cannot be used to make causal inferences
127
What are the characteristics of a measure?
reliability and validity
128
What is reliability of a measure?
the consistency of your measure; will your measure produce similar results when repeated measurements are taken under ideal circumstances
129
What are two components that any measure has?
true score, measurement error
130
What is the true score?
the real score on a variable
131
What is the measurement error?
the difference between the measured value of a variable and the true value
132
How do measures and reliability relate?
A measure with low measurement error will have high reliability, and a measure with high measurement error will have low reliability
133
A measure with low measurement error will have _____ __________.
high reliability
134
a measure with high measurement error will have ___ __________.
low reliability
135
What are the different types of reliability?
test-retest reliability
split-half reliability
inter-rater reliability
136
What is test-retest reliability?
is the extent to which a test yields consistent results over time
For this type of reliability participants are given the same test at two different points in time to see whether the results are the same across both administrations.
Similar scores across the two administrations demonstrate reliability (consistency in the measuring device).
137
What are the problems with test-retest reliability?
Subjects may remember how they responded on the first administration and respond accordingly on the second making the measure appear reliable.
The person may "change" between the two administrations making the measure appear unreliable (e.g., assessing the test-retest reliability of a mood questionnaire and the person’s mood changes between the two administrations making the test appear unreliable).
138
What type of variables is test-retest reliability best for?
more stable psychological variables such as personality
139
WHat is split-half reliability?
an example of internal consistency reliability.
Total score on one half of the test is correlated with the total score on the other half of the test (two forms of the same test).
If the test is reliable then performance on one half of the test will be related to performance on the other half of the test.
The two halves of the test are administered at the same time so the variable being measured has no time to change.
140
What is the problem with split-half reliability?
the two forms are not equivalent
141
WHat is inter-rater reliability?
the consistency across raters (observers, researchers, etc.).
A measure has good inter-rater reliability to the extent that different raters obtain the same measurement of the same variable.
For example, two researchers counting aggressive acts made by children in the schoolyard during recess should count approximately the same number of aggressive acts if they are using the same operational definition of aggression – the consistency across the two raters is the measure of inter-rater reliability.
142
What are Correlation coefficients typically used to assess?
reliability.
143
correlation coefficient: The higher the correlation between the two administrations/halves/raters the _______ _____ _____________.
higher the reliability.
144
What is validity?
the extent to which a test measures what it is designed to measure
145
WHat is construct validity?
is the extent to which a test measures the theoretical variable (construct) that it is supposed to be measuring
146
What is a construct?
a psychological variable that is not directly observable; that has been developed to explain behaviour on the basis of some theory (e.g., intelligence, self-esteem, aggression etc.)
147
What are indicators of construct validity?
concurrent validity
convergent validity
content validity
face validity
predictive validity
discriminant (divergent) validity
148
What is the correlational method?
Determines whether, and to what extent, two variables are related to one another
*correlations are used to describe the relationships between variables and not to explain them (i.e., it is simply descriptive)
149
What is face validity?
how well the test appears to measure what it was designed to measure
150
What is concurrent validity?
the measure should be able to distinguish between groups that it should theoretically be able to distinguish between.
The scores on your measure and on the criterion are collected at the same time (criterion - the variable you want to predict).
For example, a measure of bipolar disorder should be able to distinguish between people who are diagnosed with bipolar disorder and those who are not; that is, the test should be able to predict who has bipolar disorder.
151
What is predictive validity?
the ability of a test to predict a future outcome; you compare scores on your test with the criterion measure observed at a later point in time.
For example, do SAT scores predict success in college?
The criterion is the variable you want to predict so in the example above the criterion is success in college.
152
What is convergent validity?
is the extent that the scores on your measure are related to score on other measure that measure the same construct.
For example, you would expect two separate measures of motivation to be correlated as they are both measuring the same construct
153
What is discriminant (divergent) validity??
to the extent that the measure is not related to constructs to which it should not be related to .
For example, a measure of shyness should not correlate positively with a measure of extroversion
154
What is content validity?
is the extent to which a measure reflects the specific intended domain of interest .
For example, a survey to test children’s math skills should not only include multiplication questions but should also include other mathematical functions such as addition, subtraction and division.
155
Can a measure be reliable but not valid?
yes, but if a measure is not reliable, it is not valid.
156
Can a measure be valid but not reliable?
No, if a measure is valid, it must be reliable.
157
What is a scale?
a classification scheme that describes the nature of information within the values that you assign to variables
158
The scale that you decide to use determines the _______ ________ that you can perform on the data.
statistical tests
159
What are nominal scales?
the numbers are used to refer to categories, that is, they have no numerical value.
The numbers simply refer to differences in category.
For example, categorizing major in college/university - the numbers do not indicate more or less of a major, they just categorize the majors :
1 = History
2 = English
3 = Psychology
160
What is the correlational coefficient?
A statistical measure of the extent that two variables are related to one another
-single number between -1.0 and +1.0 that says the strength and direction of the relationship between two variables
- (r) = pearson product-moment correlation coefficient
161
What is an ordinal scale?
ranking; they involve quantitative distinctions by allowing us to rank order people or objects on the variable being measured.
The number start taking on a numerical meaning.
A higher value means there is more of the variable being measured but it doesn't tell you how much more.
For example, 1st, 2nd, and 3rd in a race – here people are ranked in terms of speed but the difference between each is not necessarily equal.
162
What are interval scales?
the intervals between the numbers are equal in size.
There is no absolute zero, which indicates the absence of the variable.
You cannot make meaningful ratio statements.
For example, IQ would be measured on an interval scale
- A score of 0 on an IQ test does not indicate the absence of intelligence
- You cannot therefore say that a person who scored 140 on an IQ test is twice as intelligent as a person who scored 70
163
What is a ratio scale?
equal intervals and an absolute zero that indicates the absence of the variable.
For example, weight, time, number of responses, number of items remembered are all examples of variables measured on a ratio scale.
With these types of scales you can make ratio statements such as "a person who weighs 150 pounds weighs half as much as someone who weighs 300 pounds" .
164
What is sensitivity?
a measure is sensitive to the extent that it detects small but real differences in what is being measured
165
What are range effects?
the sensitivity of a measure can be affected by a restricted range of scores; they occur when the values of the measure have an upper or lower limit:
floor / ceiling effects
166
What are floor effects?
although differences between conditions may exist, they cannot be detected because all conditions perform near zero on the measure
167
What is the pearson product-moment correlation coefficient
The most widely used correlation coefficient
(r)
168
What are ceiling effects?
although differences between conditions may exist, they cannot be detected because all conditions perform near 100%
169
What is reactivity?
subjects may behave differently in the experimental session (compared to the real world) simply because they are in an experiment.
The behaviour, therefore, may not be representative of normal behaviour, thus affecting the validity of the findings.
For example, the subject may try to discover the hypothesis and then try to behave in ways that are either consistent or inconsistent with the hypothesis.
170
What are demand characteristics?
the aspects of the experimental situation that affect the way subjects behave, that is they demand that the participants behave in a certain way (this is why they are called demand characteristics).
For example:
The experimenter wearing a lab coat, the equipment being used, etc. would be examples of demand characteristics.
Based on those characteristics, the subject may try to discover the hypothesis and then try to behave in ways that are either consistent or inconsistent with the hypothesis.
171
What are ways of neutralizing the effect of demand characteristics?
Use partial disclosure.
Allow participants to habituate to the researcher’s presence.
In a questionnaire, use filler items (items that will not be analyzed that can take the participants attention away from the real purpose of the questionnaire).
Use naturalistic observation.
Conduct a post-experimental inquiry - find out whether subjects had guessed the true purpose of the experiment and then eliminate those who did.
172
What is experimenter bias?
any intentional or unintentional influence that the experimenter exerts on the subject in an attempt to confirm the hypothesis
173
What are ways of neutralizing the effects of experimenter bias?
run participants in all conditions at the same time.
using blind techniques.
automate your experiment (have a computer collect the data).
174
What is a single-blind procedure?
the participant is unaware of which condition they have been placed in- it eliminates the possibility of reactivity
175
What is a double-blind procedure?
the participants and researcher are unaware of which condition the participants are in - it eliminates the possibility of experimenter bias and reactivity
176
What does the double-blind procedure control for?
experimenter bias and reactivity
177
What does the single-blind procedure control for?
reactivity
178
What are the two ways r can vary?
1. The direction of the relationship can be positive or negative
2. The strength of the relationship ranges from +1 to -1
-the positive or negative sign indicates the direction of the relationship (positive or negative)
-the number indicates the strength of the relationship
*the closer to 0, the weaker the relationship
179
What is automating your experiment?
using computers, etc. to present stimuli and collect data
Such measures tend to be more accurate and less variable.
However, you may miss important details within an experimental session and these details may provided insight into behaviour.
180
What does positive or negative correlation mean the variables are doing?
A Positive Correlation is when the two variables increase or decrease together in the same direction
Negative Correlation is when one variable increases and the other decreases in the opposite direction
181
What is a unit of analysis?
the entity that you wish to be able to say something about at the end of your study
It is the ‘who’ or ‘what’ you are studying
182
What is a one-group pretest posttest design
A quasi experimental design in
which the effect of an independent variable is inferred from the pretest-posttest difference in a single group.
To illustrate, suppose you wanted to test the hypothesis that a relaxation training program will decrease cigarette smoking. If you were to use the one-group pretest-posttest design, you would select a group of people who smoke, measure their smoking rate, have them attend
relaxation training, and then measure their smoking rate again.
183
What is a one-group posttest-only design?
a quasi-experimental design that has no control group and no pretest comparison; a poor design in terms of internal validity
For example, employees in a company might participate in a four-hour information session on emergency procedures, after which they score an average of 90 percent on a knowledge test. Without any sort of comparison, it would be inappropriate to conclude that the program is successfully
educating employees. Remember, this design lacks internal validity
184
What is a non-equivalent control group design?
a quasi-experimental design in which the groups of participants in the different conditions are not equivalent (e.g., naturally occurring groups), and there is no pretest .
One example is when the participants are not randomly assigned but are instead chosen
from naturally pre-existing groups (e.g., students enrolled in the morning section for a
course versus the evening section).
185
What is a scatterplot?
A graph in which paired scores for many
subjects are plotted as single points to reveal the direction and strength of their correlation
-The more spread out the dots appear, the weaker the correlation.
-The closer together the dots appear, the stronger the correlation.
*Its a visualized relationship between two variables
186
What is a non-equivalent control group pretest-posttest design?
a quasi-experimental design in which non-equivalent groups are used, but a pretest allows assessment of equivalency and pretest-posttest changes.
This is not a true experimental design because assignment to groups is not random, and so
the two groups may be different at the beginning. However, we now have the advantage of knowing pretest scores, so we can see whether the groups were the same on the pretest, at least. Moreover, even if the groups are not equivalent, we can look at changes in scores from
the pretest to the posttest.
187
What direction does the line of a scatterplot go for positive and negative correlations?
Positive: line goes up
negative: line goes down
188
What is an interrupted time series design?
A quasi experimental design in which a treatment is investigated by examining a series of measurements made over an extended time period, both before and after the treatment is introduced.
189
What is a control series design?
an extension of the interrupted time series quasi-experimental design in which there is a comparison or a control group
190
How do you create a scatterplot?
* you place one variable on the x-axis
* you place the other on the y-axis
* and then you take each participant’s pair of scores and make a single point on the graph
191
Why use correlations?
1. Exploratory research
2. They can be done when ethics prohibit experiments
3. They may be able to challenge a theory that holds that two variables should not be correlated
4. Making predictions:
e.g., if you know that SAT scores are positively correlated with college GPA, then you can use a student’s SAT score to predict (within limits) the GPA the student is likely to achieve
192
What is a predictor variable?
The variable used to predict
193
What is the criterion variable?
The variable being predicted
**criterion is a word of the day
194
Which axis does the predictor and criterion variables each go on?
Predictor goes on the x-axis
Criterion goes on the y-axis
195
Why are the conclusions drawn from correlations limited?
You cannot use them to make causation statements, only to describe relationships between variables
196
What is the problem of directionality? What's the third variable problem?
Even if a causal relationship existed between the two variables, the direction of causality may be difficult to determine:
We cannot make causation statements because if two variables (X and Y) are correlated, it may be that:
-X has a causal influence on Y
or
-Y has a causal influence on X
or
Third variable problem: some third variable, Z, has a causal influence on both X and Y
197
Can r detect curvilinear relationships?
No
-Correlations are only designed to detect linear relationships.
-If the relationship is curvilinear the correlational coefficient will not indicate that presence of a relationship.
198
Describe hypothesis testing with correlation coefficients
When testing hypotheses, you always set up two hypotheses: null and alternative
-if the results are statistically significant then you reject the null and provide support for the alternative hypothesis
199
What is the null and alternative hypothesis?
-Null hypothesis - the assertion that the correlation between the two variables in the population is zero (i.e., that there is no
correlation between the two variables)
-Alternative hypothesis - the assertion that there is a relationship between the two variables in the population (that is, the value
you obtained in your correlation did not occur by chance)
200
What effect does sample size have with significance?
The larger the sample, the more likely the results will be significant
201
What is the level of significance?
Predetermined probability below which the results are called statistically significant
-typically set at .05; that is, if there is less than a 5% chance that the results are due to random error the experimenter will
reject the null hypothesis, which provides support for the alternative hypothesis
202
What are descriptive and inferential stats?
Descriptive: used merely to describe a set of data
Inferential: Allows us to make inferences about populations based on results obtained from samples drawn from those populations
- This is how we determine statistical significance
203
What are three descriptive stats?
sample size, measures of central tendency, measures of variability
204
What is sample size?
(n) - the number of observations that make up a data set
205
What are measures of central tendency? What are the 2 subparts?
A single value that describes the typical or central score in a data set
1. Mode: the most common value in a data set
2. Median: the middle score in a data set when the data are arranged in order
206
How do you find the median (md)? (measures of central tendancy)
- Arrange all the scores in numerical order (if a number is repeated, be sure to include all
instances)
- Find the score that corresponds to the middle of the dataset, so that half the scores are above and half the scores are below.
*If there is an odd number of scores in the dataset, the center score is the median
*If there is an even number of scores in the dataset, the median will be the average of the two middle scores.
207
What's the mean? (measures of central tendency)
The average score in a dataset obtained by adding all the scores in the dataset and then dividing by the number of scores (n)
*the symbol for mean is an X with a line over it
*we need to know the formula for this!!*
208
How do you decide which measure of central tendency to use?
* If the data are on a nominal scale you need to use need to use the mode as you measure of central tendency
* If the data are on an ordinal scale then use the mode or the median (you can't use the mean because it is
sensitive to the distance between scores)
* If the data are Interval or Ratio in nature, use the median or mean, depending on the shape of the distribution:
- If the distribution is normal use the mean because it provides the most information
- If the distribution is skewed use the median as it is not influenced by the outliers as much as the mean
sorry lol
209
What are measures of variability? What are the two things involved?
A single value that describes the spread in the data set
Range, variance and standard deviation
210
What is range? How is it calculated?
* A single score that represents the difference
between the highest and lowest score in a dataset
* It is calculated by subtracting the lowest score from the highest score
* It depends entirely on the highest and lowest
scores and tells you nothing about how spread out the rest of the scores are in the dataset
* The range is therefore not very informative
211
What is variance and standard deviation?
Both measures of the average distance of each score from the mean
-both provide an indication of how spread out the scores are relative to the mean
-they are mathematically related in that the standard deviation is just the square root of the variance
**we need to memorize the formula for each of these**
212
What are frequency distributions?
* To get a visual of a dataset, researchers plot it in a frequency distribution.
* They plot all values of the DV against their frequency of occurrence, i.e., how often each one occurs.
* The DV is placed on the X axis
* The frequency of occurrence is placed on the y axis
213
How do you create a frequency distribution
1. Put the DV is on the X axis
2. Put frequency of occurrence on the y axis
3. Then each value is place on the graph
214
I cant turn this into a card, but we need to be able to identify different shapes of frequency distributions so just go to the stats lecture and look at the examples
:')
215
What is normal distribution?
- the shape of the distribution determines which measures of central tendency you should use
- If the distribution is normal (symmetrical and
unimodal) the mean, median, and mode will all be approximately the same value so use the mean as your measure of central tendency as it provides the most amount of information.
- a symmetrical distribution of scores
in which the mode, median, and mean are the same value (it is often called the ‘bell curve’)
216
What is a population parameter?
Any numerical quantity (e.g., mean) that characterizes a given population
Examples:
mu (the population mean) – μ
sigma (the population standard deviation) - σ
217
What are z scores?
We change our score into a z score so
that we can refer to the z score tables (the standard normal distribution)
Specifically, we have to convert our data to what are called z scores. To do this, we take the scores, subtract the population mean (mu = μ) and then divide by the population standard deviation
This transformation is called z.
218
What is percentile score?
The percentage of people or observations that fall below a given score in a normal distribution
219
What are inferential statistics?
Making inferences about populations based on results obtained from samples drawn from those populations
220
Do inferential stats use the null and alternative hypothesis?
Yes
-The null hypothesis states that the results are
due to chance: that there is no real difference
between the two groups (i.e., the IV had no
effect OR the control and experimental groups
will not differ in the amount of sleep)
-The alternative hypothesis states that the
result are not due to chance: that there is a
real difference between the two groups (i.e.,
the IV did have an effect OR the control and
experimental groups will differ in the amount
of sleep)
221
Word of the day: Alpha - define it
* The level of statistical significance set in an experiment (usually set at 0.05 or 5%)
* It is also known as the Type I Error Rate
222
What is sampling error?
The variability in a statistic from one sample to another due to chance, for example, the mean, sd, etc. will vary across samples, even if the samples are being drawn from the same population
223
What is sampling distribution of the mean?
Is the distribution of the means calculated from an infinite number of random samples drawn from a specified population
224
What is alpha in stats?
* alpha is the critical cutoff that we set ahead of time (typically set at 5%)
* this means that if our results show that there is less than a 5% chance that our obtained sample mean occurred in this sampling distribution, we will reject the null and conclude that it was sampled from a different population with a different mean
225
What are type 1 errors
Rejecting the null hypothesis when it should have been accepted
- 5% of the time we will reject the null when it is true, this is called a type I error and it is equivalent to alpha
- So 5% of the time, we will reject the null when we should have accepted it
*type 1 errors are directly tied to the p value set in an experiment
226
What are type 2 errors?
(beta) - accepting the null when it is false
-Decreasing the type I error rate increases the type II error rate
227
What is power?
The ability to detect the effects of the IV if they actually exist
228
What is the difference between one tailed and two tailed tests?
One-tailed tests allow for the possibility of an effect in one direction. Two-tailed tests test for the possibility of an effect in two directions—positive and negative.
229
Qualitative vs quantitative
-With qualitative research the aim is to describe verbally and in detail the behaviour being observed; here each observation is a word, object, description, or idea.
-With quantitative research the aim is to reduce behaviour to a quantity; here each observation is reduced to a number
-With quantitative research the interest is in
measurement and with qualitative research it is in description.
230
What is also known as predictive validity?
Criterion validity
231
What is also known as discriminant validity?
divergent validity
