1
Q
descriptive research questions
A
Questions that examine and describe what already exists (e.g., how many people are involved in car accidents each year).
2
Q
causal research questions
A
Questions that attempt to determine cause-and-effect relationships between variables.
3
Q
nondirectional hypothesis
A
A hypothesis that states a relationship exists but does not specify the direction of the relationship.
4
Q
directional hypothesis
A
A hypothesis that specifies the expected direction of the relationship between variables.
5
Q
null hypothesis typically rejected
A
When the p value is less than or equal to the significance level (α).
6
Q
p value
A
is a statistical measure indicating the likelihood of observing your data, or something more extreme, if the null hypothesis (no effect or no difference) were true, helping determine statistical significance
7
Q
p value indicate the probability that the null hypothesis is true
A
No.
8
Q
Type I error (α)
A
Rejecting a null hypothesis that is actually true (false positive).
9
Q
Type II error (β)
A
Failing to reject a null hypothesis that is actually false (false negative).
10
Q
alpha (α)
A
The probability of committing a Type I error.
11
Q
Type II error when alpha is decreased
A
The probability of committing a Type II error increases.
12
Q
statistical power
A
The likelihood of detecting a significant effect when one truly exists (1 − β).
13
Q
power increases when…
A
Increasing alpha, increasing sample size, increasing effect size, minimizing error, using a one-tailed test, or using parametric statistics.
14
Q
probability sampling
A
Sampling from a known population where each member has a known chance of selection.
15
Q
nonprobability sampling
A
Sampling that does not involve random selection and often uses convenience samples.
16
Q
simple random sampling
A
Every member of the population has an equal chance of being selected.
17
Q
systematic sampling
A
Selecting every nth individual from a population list.
18
Q
stratified random sampling
A
Dividing a population into subgroups and randomly sampling from each subgroup.
19
Q
cluster sampling
A
a cost-effective survey method where a large population is divided into naturally occurring groups (clusters), like neighborhoods or schools, and then researchers randomly select a few of these clusters to study entirely, rather than surveying individuals from the whole population
20
Q
multistage sampling
A
Sampling that occurs in multiple random stages (e.g., districts → schools → classes).
21
Q
convenience sampling
A
Selecting participants who are easily accessible.
a non-probability research method where participants are chosen because they are easy and quick for the researcher to access, such as surveying people at a mall or online friends, offering speed and low cost but suffering from significant bias, as the sample often doesn’t represent the broader population.
22
Q
purposeful sampling
A
a non-probability research technique where researchers strategically select participants or cases based on specific characteristics, knowledge, or experiences relevant to the study’s goals, aiming for rich, in-depth understanding rather than broad generalization, common in qualitative research to explore phenomena like cancer survivors’ experiences or teachers’ strategies with special needs students
23
Q
quota sampling
A
A non-probability survey method where researchers select participants to match population proportions for specific traits (like age, gender) to ensure representation, but without random selection
24
Q
internal validity
A
The degree to which changes in the dependent variable are caused by the independent variable.
25
experimenter effects
Biases where the researcher influences participant responses.
26
halo effect
When initial positive perceptions of a participant influence judgments of other traits.
27
Hawthorne effect
When participants change behavior due to awareness of being observed.
28
subject effects
Changes in participant behavior due to their role in the study.
29
demand characteristics
Cues that influence participants to behave in certain ways during a study.
30
population external validity
The degree to which results can be generalized to other populations.
31
ecological external validity
The degree to which results can be generalized to other settings or conditions.
32
quantitative research
Research that examines relationships between variables measured numerically.
33
examples of quantitative research
Wait time measurements, experimental drug studies, and survey-based research.
34
qualitative research
Research that explores how or why phenomena occur using non-numerical data.
35
types of data used in qualitative research
Interviews, field notes, pictures, videos, and artifacts.
36
mixed-methods research
Research that combines quantitative and qualitative approaches.
37
concurrent mixed-methods design
Collecting qualitative and quantitative data at the same time (triangulation).
38
sequential mixed-methods design
Collecting one type of data first, followed by the other.
39
single-subject research designs (SSRDs)
evaluating interventions by intensely studying one subject (person, family, class) over time, repeatedly measuring their behavior (dependent variable) across different conditions (baseline, intervention) to establish cause-effect, often using visual analysis of data (level, trend, latency) rather than complex statistics, common in behavior analysis, special ed, and clinical fields to test treatments on individuals before large-scale studies
40
descriptive research
Research used to describe a phenomenon without introducing an intervention or treatment.
41
main limitation of descriptive research
It can describe what exists and how often it occurs, but it cannot explain why it occurs.
42
descriptive research often conducted first
Data must be described and summarized before examining relationships between variables.
43
example of descriptive research
Showing that children watch an average of three hours of television per day.
44
longitudinal research
Research that collects data from the same group over an extended period of time.
45
purpose of longitudinal research
To track patterns or developmental changes in behavior over time.
46
major limitations of longitudinal research
High cost, cohort effects, and participant attrition (mortality).
47
cohort effect
Effects caused by shared experiences among individuals in the same group.
48
cross-sectional research
Research that examines different groups or cohorts at a single point in time.
49
example of cross-sectional research
Comparing TV viewing effects on test scores across different age groups at the same time.
50
key limitation of cross-sectional research
Differences observed may not reflect true developmental change.
51
conclusions in cross-sectional research only be inferred because…
Because the same individuals are not studied over time.
52
survey research
A method of collecting quantitative and qualitative data using questions administered to participants.
53
forms of surveys
Written questionnaires, verbal questionnaires, interviews, or written statements.
54
example of large-scale survey research
The U.S. Census.
55
importance of survey design
Poor wording, format, or inappropriate difficulty can make results unreliable.
56
participant factors in survey research that affect accuracy
Reading level, comprehension ability, and understanding of questions.
57
action research
Research conducted by professionals to improve their own practice or organizational effectiveness.
58
defining characteristic of action research
It is site-specific and applied to a particular school, agency, or organization.
59
goal of action research
To test new approaches and reflect on practice to enhance effectiveness.
60
example of action research
Conducting a needs assessment among middle-school students.
61
another example of action research
Interviewing adolescents about experiences in a residential facility.
62
third example of action research
Presenting test scores to justify an intervention.
63
nonexperimental research designs
Research designs that are exploratory and descriptive, with no intervention or manipulation of variables.
64
primary goal of nonexperimental research designs
To observe and describe the properties of a variable.
65
Do nonexperimental research designs involve manipulation of variables?
No.
66
experimental research designs
Designs that involve an intervention where the researcher manipulates conditions or variables.
67
primary goal of experimental research
To assess cause-and-effect relationships among variables.
68
required for most experimental research designs
Random assignment of participants.
69
types of group comparisons in experimental designs
Single-group designs, treatment vs. control groups, or comparisons between two treatment groups.
70
Are single-subject research designs (SSRDs) qualitative or quantitative?
Primarily quantitative, though they may include qualitative components.
71
What do SSRDs measure?
Behavioral and/or attitudinal changes over time in one or a few individuals.
72
Why are SSRDs considered experimental in nature?
Because they systematically measure changes across time, often in response to an intervention.
73
What is a nonexperimental research design?
A design that lacks manipulation of independent variables and does not require random assignment.
74
purpose of nonexperimental research designs
To observe, describe, compare, or examine relationships among variables without intervention.
75
four types of nonexperimental research designs
Descriptive, comparative, correlational, and ex post facto designs.
76
descriptive research design
A design that thoroughly describes a variable at one time or over time.
77
two main forms of descriptive design
Simple descriptive design and longitudinal design.
78
simple descriptive design
A one-shot survey describing a single variable at one point in time.
79
example of a simple descriptive design
Studying the average number of counseling sessions couples attend in a city.
80
cross-sectional design
A special type of simple descriptive design that studies different groups at the same time.
81
example of a cross-sectional design
Comparing alumni donations from graduates 1, 5, 10, and 20 years post-graduation.
82
longitudinal design
A design that studies variables over time.
83
three types of longitudinal designs
Trend, cohort, and panel studies.
84
trend study
Assessing the general population over time using different participants at each data collection.
85
cohort study
Studying the same population over time.
86
panel study
Studying the same individuals over time.
87
comparative research design
A design that examines group differences on a variable.
88
major limitation of comparative designs
They cannot establish causation.
89
example of a comparative design
Examining gender differences in math achievement scores.
90
correlational research design
A design that describes the relationship between two variables.
91
correlation coefficient
The strength and direction of a relationship between two variables.
92
shared variance
The amount of overlap between two variables, derived from the correlation coefficient.
93
shared variance (coefficient of determination) calculation
By squaring the correlation coefficient (r²).
94
If r = .50, what is the shared variance?
.25, or 25%.
95
ex post facto research design
A nonexperimental design that examines possible cause-and-effect relationships after data have been collected.
96
another name for ex post facto design
Causal-comparative design.
97
Why does ex post facto research resemble experimental research?
Because it examines potential cause-and-effect relationships.
98
Why can causation not be confirmed in ex post facto research?
Independent variables cannot be manipulated and randomization is not possible.
99
What is a within-subject design?
A design that assesses changes within the same participants as they experience one or more interventions.
100
What is another term often associated with within-subject designs?
Repeated-measures design.
101
What does a within-subject design typically measure?
Changes in the dependent variable before and after an intervention.
102
Can within-subject designs involve more than one intervention?
Yes, multiple interventions can be introduced sequentially and measured over time.
103
What is a between-groups design?
A design that compares the effects of a treatment or intervention across two or more separate groups.
104
How are participants assigned in a between-groups design?
Each group consists of a separate sample.
105
What role can groups play in a between-groups design?
Groups may serve as control groups or receive different treatments.
106
What is a split-plot design?
A design that assesses a general intervention applied to a whole group and additional treatments applied to subgroups.
107
Why are split-plot designs useful in counseling research?
They allow researchers to evaluate both overall interventions and specific components simultaneously.
108
Give an example of a split-plot design.
Studying a mentoring club overall, then comparing résumé writing, job shadowing, and interviewing skills subgroups.
109
What is a pre-experimental design?
A design that does not use random assignment and lacks control for internal validity threats.
110
Why are pre-experimental designs not considered true experiments?
Because they do not adequately control for threats to internal validity.
111
How many types of pre-experimental designs are there?
Three.
112
What is a one-group posttest-only design?
A single group receives an intervention and is measured only after the intervention.
113
What is a one-group pretest–posttest design?
A single group is measured before and after an intervention.
114
What is a nonequivalent groups posttest-only design?
A design with two nonrandom groups where one receives treatment and both are measured after the intervention.
115
What is a true experimental design?
A design that uses random assignment and includes at least two groups for comparison.
116
What distinguishes true experimental designs from quasi-experimental designs?
Random assignment.
117
Why are true experimental designs considered the gold standard?
They provide the strongest control over internal validity and allow causal conclusions.
118
What is a randomized pretest–posttest control group design?
Participants are randomly assigned to treatment or control groups and measured before and after intervention.
119
What is a randomized pretest–posttest comparison group design?
Participants are randomly assigned to groups that receive different interventions, with pretests and posttests.
120
What is a randomized posttest-only control group design?
Participants are randomly assigned to treatment or control groups and measured only after intervention.
121
What is a randomized posttest-only comparison group design?
Participants are randomly assigned to two or more treatment groups with no control group and no pretest.
122
What is the Solomon four-group design?
A comprehensive true experimental design combining pretest–posttest and posttest-only designs.
123
What is the purpose of the Solomon four-group design?
To evaluate both treatment effects and pretest effects.
124
What does Group 1 receive in the Solomon four-group design?
Pretest, intervention, and posttest.
125
What does Group 2 receive in the Solomon four-group design?
Pretest and posttest, but no intervention.
126
What does Group 3 receive in the Solomon four-group design?
Intervention and posttest, but no pretest.
127
What does Group 4 receive in the Solomon four-group design?
Posttest only (no pretest or intervention).
128
What is a quasi-experimental design?
A design used when random assignment is not possible or appropriate.
129
When are quasi-experimental designs commonly used?
With naturally occurring or nested groups (e.g., classrooms or counseling groups).
130
What is a nonequivalent groups pretest–posttest design?
Intact groups are given a pretest, one or more groups receive treatment, and all groups receive a posttest.
131
What is a time series design?
A design involving repeated measurements before and after an intervention.
132
What is an interrupted time series design?
A time series design where an intervention clearly interrupts the baseline pattern.
133
What is required for time series designs to be valid?
Equal time intervals, consistent measurement procedures, and a clearly defined intervention.
134
Single-Subject Research Designs (SSRDs)
Designs that allow repeated measurement of a target behavior over time for an individual or a small group.
135
Why are SSRDs useful for counselors?
They provide concrete evidence of treatment effectiveness for specific clients.
136
What does “A” represent in SSRDs?
Baseline phase (no treatment).
137
What does “B” represent in SSRDs?
Treatment or intervention phase.
138
What are the three main types of SSRDs?
Within-series designs, between-series designs, and multiple-baseline designs.
139
What is a within-series design?
A design that examines the effectiveness of one intervention over time.
140
What is an A–B design?
A baseline phase followed by a treatment phase; similar to a posttest-only design.
141
What is an A–B–C design?
A design that examines the interaction among multiple treatment components.
142
What is a changing-criterion design?
A design where success criteria become progressively more restrictive to assess treatment effectiveness.
143
What is a parametric design?
A design that compares different treatment intensities or variations across phases.
144
What is a between-series design?
A design that compares the effectiveness of two or more interventions for a single variable.
145
What is a multiple-baseline design?
A design that measures a target behavior across individuals, settings, or behaviors with staggered intervention introduction.
146
Why are interventions staggered in multiple-baseline designs?
To demonstrate that behavior change occurs due to the intervention, not chance or time.
147
What are the three ways multiple-baseline designs can be applied?
Across individuals, across environments, or across behaviors.
148
Give an example of a multiple-baseline across individuals design.
Introducing the same intervention at different times for Juan, Billy, and Tamika.
149
Give an example of a multiple-baseline across environments design.
Applying an intervention sequentially at home, school, and sports practice.
150
Give an example of a multiple-baseline across behaviors design.
Sequentially targeting raising hand, staying seated, and keeping hands to self.
151
Frequency polygon
The frequency polygon is a line graph of the frequency distribution.
The X-axis typically indicates the possible values, and the Y-axis typically
represents the frequency count for each of those values. A frequency
polygon is used to visually display data that are ordinal, interval, or ratio.
152
Histogram
Displays how often values fall within numerical ranges
Used for continuous variables (e.g., age, test scores, income)
153
Bar graph
Displays frequencies or values for distinct categories
Used for categorical variables (e.g., gender, diagnosis type)
154
Outlier
An outlier is an extreme data point
that distorts the mean (i.e., it inflates or deflates the typical score).
155
Interquartile range
When dealing with outliers, and thus potentially skewed distributions, the interquartile range
may be a more accurate estimate of variability, as it eliminates the top and bottom quartiles and provides a range
around the median score (i.e., the middle 50%).
To calculate the interquartile range, divide the range of scores into four equal parts.
Then, subtract the score that is one quarter from the bottom from the
score that is three quarters from the bottom. Divide this number by 2,
and this becomes the plus or minus range for the interquartile range.
156
Standard deviation
a statistical measure showing how spread out data points are from their average (mean)
The standard deviation of a known population is indicated by σ
157
Sum of squares
representing the total of squared differences from a value (often the mean) to measure data spread (variability)
158
Varience
showing how spread out numbers in a dataset are from their average (mean), calculated as the average of the squared differences from the mean
159
Kurtosis
The degree of peakedness of a distribution. Distributions can be mesokurtic (normal curve), leptokurtic (tall and thin), and platykurtic (flat and wide)
160
Inferential statistics
Statistical procedures that are used to draw inferences about a population
from a sample.
161
Degrees of freedom
the number of values in a dataset that are free to vary when estimating a parameter
The count of values that can change freely in a calculation without violating any constraints.
N-1
162
Valence
The + or - in correlations
Can be positive correlation or negative correlation
163
What is the possible range of correlation values?
From −1.00 to +1.00.
164
What does the Pearson product-moment correlation coefficient measure?
The strength and direction of a linear relationship between two continuous variables.
165
What is Pearson correlation commonly called?
Pearson r.
166
What does a correlation of +1.00 indicate?
A perfect positive relationship.
167
What does a correlation of −1.00 indicate?
A perfect negative relationship.
168
What does a correlation close to 0 indicate?
Little to no linear relationship.
169
When is Pearson r used?
When both variables are continuous and normally distributed.
170
What is Spearman r used for?
Correlating rank-order (ordinal) variables.
171
What is a biserial correlation coefficient?
A correlation between one continuous variable and one artificially dichotomous (dummy-coded) variable.
172
What is a point biserial correlation coefficient?
A correlation between one continuous variable and one true dichotomous variable.
173
spurious correlation
A correlation that overrepresents or underrepresents the true relationship between variables.
174
What causes overestimation in correlations?
When variables overlap or when a third unmeasured variable affects both variables.
175
Give an example of a third-variable problem.
Age influencing both reading achievement and body weight.
176
attenuation in correlation
A weakened correlation caused by unreliable measurement.
177
How does low reliability affect correlation strength?
It lowers the observed relationship between variables.
178
restriction of range
When a sample is not representative of the population, limiting variability.
179
How does restriction of range affect correlations?
It results in inaccurate or misleading relationships.
180
Can both overly homogeneous and overly heterogeneous samples distort correlations?
Yes.
181
Why shouldn’t correlations be interpreted as percentages?
Because the correlation coefficient itself is not a percentage.
182
coefficient of determination
The percent of variance shared between two variables.
183
How is the coefficient of determination calculated?
By squaring the correlation coefficient (r²).
184
If r = .40, what is the shared variance?
.16, or 16%.
185
What three things commonly distort correlations on exams?
Third variables, unreliable measures, and restriction of range.
186
What are prediction studies?
Extensions of correlational studies used to predict outcomes, also known as regression studies.
187
What is a key limitation of prediction (regression) studies?
They predict outcomes but do not explain causation.
188
What is a predictor variable in regression?
The independent variable used to predict outcomes.
189
What is a criterion variable in regression?
The dependent variable being predicted.
190
What is bivariate regression?
Regression using one predictor variable to predict one criterion variable.
191
What is multiple regression?
Regression using two or more predictor variables to predict a criterion variable.
192
What are beta weights?
Statistical weights that show the relative contribution of each predictor in multiple regression.
193
How does adding predictor variables affect prediction accuracy?
Generally, more predictors lead to stronger prediction.
194
What is logistic regression?
Regression used when the dependent variable is dichotomous.
195
Can regression studies establish cause-and-effect relationships?
No—they only predict outcomes.
196
What type of regression is used to predict a yes/no outcome?
Logistic regression.
197
What are parametric statistics?
Statistical tests used when assumptions are met (e.g., normality, homogeneity of variance).
198
What does a t-test compare?
Two means for one dependent variable.
199
What is an independent t-test?
A t-test comparing two independent groups on one dependent variable.
200
Give an example of an independent t-test.
Gender differences in achievement scores.
201
What is a dependent t-test (repeated measures t-test)?
A t-test comparing the same group tested twice or matched/paired groups.
202
Give an example of a dependent t-test.
Pretest vs. posttest scores for the same students.
203
What statistic does a t-test produce?
A t ratio.
204
What is ANOVA used for?
Comparing three or more group means for one independent variable.
205
Why is ANOVA preferred over multiple t-tests?
To reduce Type I error.
206
What statistic does ANOVA produce?
An F ratio.
207
What does a significant F ratio indicate?
At least two group means differ significantly.
208
What is post hoc analysis?
Tests used after a significant ANOVA to identify which groups differ.
209
When is a factorial ANOVA used?
When there are two or more independent variables.
210
What two effects does a factorial ANOVA examine?
Main effects and interaction effects.
211
What is an interaction effect?
When the effect of one independent variable depends on another.
212
Give an example of an interaction effect.
CBT works better for males, while IPT works better for females.
213
What is ANCOVA?
ANOVA that controls for a covariate (a variable statistically adjusted).
A covariate is a measurable factor that "co-varies" with the outcome but isn't the main thing being tested, like humidity affecting paint drying time when you're really studying different paint types.
214
What is a covariate?
A variable whose influence is removed from the analysis.
Think of it as a measurable factor that "co-varies" with the outcome but isn't the main thing being tested, like humidity affecting paint drying time when you're really studying different paint types.
215
Give an example of ANCOVA.
Examining income and work satisfaction while controlling for gender.
216
When should you use factorial ANOVA instead of ANCOVA?
When the covariate is also a primary independent variable of interest.
217
What is MANOVA?
ANOVA with multiple dependent variables.
218
What is MANCOVA?
ANCOVA with multiple dependent variables.
219
Which test compares 2 groups?
t-test
220
Which test compares 3+ groups?
ANOVA
221
Which test controls for another variable?
ANCOVA
222
Which tests involve multiple dependent variables?
MANOVA / MANCOVA
223
What are nonparametric statistics?
Statistical tests used when few assumptions can be made about the population distribution.
224
When are nonparametric statistics recommended?
When data are nominal or ordinal, or when interval/ratio data are not normally distributed (skewed).
225
What is the chi-square test used for?
Examining the relationship between two or more categorical (nominal) variables.
226
What is required for chi-square test scores?
All observations must be independent.
227
What does a chi-square test compare?
Observed frequencies to expected frequencies.
228
Give an example of a chi-square test.
Relationship between decision to terminate counseling (yes/no) and gender of counselor.
229
What is the Mann–Whitney U test analogous to?
An independent t-test.
230
What type of data does the Mann–Whitney U test use?
Ordinal data (ranked data).
231
What does the Mann–Whitney U test compare?
Ranks from two independent groups.
232
Give an example of a Mann–Whitney U test.
Comparing education aspirations across different grade levels.
233
What is the Kolmogorov–Smirnov Z test similar to?
The Mann–Whitney U test.
234
When is the Kolmogorov–Smirnov Z test preferred?
When sample sizes are smaller than 25.
235
What is the Kruskal–Wallis test analogous to?
A one-way ANOVA.
236
When is the Kruskal–Wallis test used?
When comparing three or more independent groups using ordinal data.
237
What is the Wilcoxon signed-ranks test equivalent to?
A dependent (paired) t-test.
238
What does the Wilcoxon signed-ranks test measure?
Amount and direction of change between paired scores.
239
Give an example of a Wilcoxon signed-ranks test.
Pre- vs. post-training competency ratings.
240
What is Friedman’s rank test similar to?
Wilcoxon’s signed-ranks test.
241
When is Friedman’s rank test used?
For repeated measures with more than two groups.
242
Match the nonparametric test to its parametric equivalent.
• Independent t-test → Mann–Whitney U
• Dependent t-test → Wilcoxon signed-ranks
• ANOVA → Kruskal–Wallis
• Repeated-measures ANOVA → Friedman’s rank
• Categorical variables → Chi-square
NCE
flashcards
Decks in class (14)
# Cards
