Research and Statistics (78) Flashcards
A correlation describes a ____________ between 2 variables
relationship
correlation describes a relationship between
two variables
a factor is also known as
an independent variable
AB Design
a pretreatment baseline measure is performed
the treatment is applied
measurement is taken after the treatment is applied
AB relies on comparing the treatment effect on either a single participant or group of participants.
ABAB Design
Measure a baseline (the first A)
the treatment is applied
Measurement is taken (the first B)
return to baseline or the withdrawal of treatment (the second A) is performed
The re-introduction of treatment is applied (The second B )
ANOVA
ANalysis Of VAriance
used with more than two groups or samples
Results of ANOVA tests are called f-ratios
Used with parametric data
Used with interval or ratio data
Are field studies considered true experimental or quasi-experimental?
Quasi-experimental
Are lab studies considered true experimental or quasi-experimental?
True experimental
Behavioral observations
Consider what happened before a target behavior, the behavior itself and its consequences.
Best fitting line is also known as
Regression line
Between Groups Design
Participants who receive the treatment are in one group
different levels of the independent variable are given to different groups of participants
one independent variable is given to one group and not to another group
the effect of the dependent variable is compared to each group receiving and not receiving the treatment
Case study vs experiment
one person is studied; not as controlled or objective as experiments.
Cluster Sampling
The population is divided into groups from which the researcher randomly selects an equal number of participants
Each group should represent the larger population from which it was selected
Coefficient of determination
The square of the correlation coefficient
an indicator of the percent of variance between variable
indicates how much the independent variable determines what the dependent variable will be
Confidence Interval
A range of values of a normal distribution of sample means which the population is expected to fall within
Confidence Level
The probability value of the confidence interval given as a percentage
Confidence Limit
the values of the lower end and the upper end of the range of the confidence interval
control group
independent variable is not manipulated in this group
correlation coefficient
ranges from -1 to 1, has both direction and magnitude
correlational method
research that examines relationships among factors does not determine cause and effect
dependent variable
variable measured in an experiment
double-blind study
experimenter and subjects do not know who is in control or experiment groups
experimental group
receives manipulation of the independent variable
external validity
ability to generalize results to groups outside the research study
homoscedasticity
the assumption that in multiple groups of data each group will have a normal distribution of data
how do you find the T-score when you are given the Z-score and standard deviation?
- Multiply the Z-score by the standard deviation
- Add the result to the mean
In a scattergram, the x-axis represents the:
independent variable.
In a scattergram, the y-axis represents the:
dependent variable
Incidence
Number of new cases of a disorder during a certain time period
Independent variable
Maniipulated in an experiment
internal validity
manipulation of the independent variable caused the effects
Leptokurtic
High peek
small range
scores are clustered together
not much variance in scores
Level of significance
A point or the alpha level at which the researcher decides, usually p = .05 that the data in one group are significantly different than the data gathered from another group
in other words, using p = .05, we can be 95% certain that our results are not due to chance and a 5% chance that they are.
Linear
linear means that given scattered points on a graph it is possible to draw a line through the scattered points on the graph that would best represent their distribution
Magnitude
Strenght of correlation, expressed as a numerical value of “r”
Mean
The average of all the scores
Adds up all the scores and divide by the number of scores
Median
place scores in order from lowest to highest. The score that lies in the middle of all scores is the median. if there is an even number of scores, take the two scores closest to the middle, add them together and divide by two to get the median.
Mesokurtic
normal bell-shaped distribution
Mode
most frequent occurring score
Multiple baseline design
improves generalizability by researching treatment that begins at different times in different groups in different settings
negative “r”
inverse relationship between variables
non=experimental research
descriptive research
correlational research
nothing is manipulated
associations between variables are determined by correlational techniques
takes less time and effort
explores information about variables that already exist to find relationships
used to predict outcomes
does not establish causal relationships between independent and dependent variables
Non-linear
non-linear means that given scattered points on a graph it is NOT possible to draw a line through the scattered points on the graph that would best represent their distribution
Null Hypothesis
the hypothesis that there is no difference between group A and group B
A and B are same
One-factor ANOVA
the researcher is considering only one independent variable
Person r is also known as
Person product moment correlation coefficient
Pearson r is used when the scores are either ________ or ________
Interval
Ratio
Percentile
a percentile tells what percent of the group scored the same or below a specific score
EX. if you scored in the 90th percentile you scored better than 90 percent of your peers, 10 percent scored better than you did
Platykurtic
curve is flat and low
range of scores is large
scores are all over the place
Prevalence
Statistics involving the presence of a disorder among a certain population at a certain time
Random assignment of subjects
Control group and experimental group should be as similar as possible
Reliability
Test gives consistent results over time
Scientific statements
Testable, based on observations, linked to measurable outcomes
Simple random sampling
Every participant in a given population has an equal chance of being included in the sample
The selection of one participant does not affect the selection fo another participant
Spearman r is also known as
Rank correlation coefficient
Spearman r is used when the scores are given as _________ scaled data.
Ordinal
Standard Error of Estimate
The error in any prediction
Standardization
similar procedures are used each time test is administered
statistical significance
probability that a relationship happened by chance
True of False
A negative correlation can be more significant than a positive correlation
True
True or False
Correlations can be greater than 1
False
True or False
Correlations can be linear or nonlinear
True
True or False
Correlations can be negative or positive
True
T-test
Used to determine is there a significant difference between the means of two groups of scores
there can only be two groups
the result of the test is known as the t ratio
if the t value is above the level of significance, reject the null hypothesis
used with interval or ratio data
used with parametric data
Type I error
You have rejected the null hypothesis
what you are saying is that there is a difference between group A and group B because of some independent variable.
But the reason this is an error is that in reality there is no difference between groups A and B because of the independent variable
you have discovered nothing but you are saying you have discovered something
this is the worst mistake you can make. You are saying there is a relationship but there is no relationship.
These types of errors can literally kill someone.
Type II error
You have accepted the null hypothesis
You are saying there is NO difference between groups A and B because of some independent variable but in reality there IS a difference.
You have saying you have discovered something when you have NOT
you are saying there is no relationship but there is a relationship
this type of error is like missing a shoe.
Validity
test measures what it says it measures
What are some methods of non-experimental research?
case studies
archival research
observational study
survey
what are three forms of quasi-experimental research?
field experiments
developmental research
ex-post factor research
what are three types of non-parametric tests of the null hypothesis?
chi square
mann-whitney y test
wilcoxon matched pairs signed rank test
what are two parametric tests of the null hypothesis?
T-test
ANOVA
What is necessary for inferential research?
Experimental conditions controlled
more than one level of independent variables
subjects are randomly assigned to groups
when and on what type of data do you use non=parametric tests?
when the population distribution is NOT normal
the type of data is ordinal or nominal
when and on what type of data do you use parametric tests?
when the population distribution is normal
the data is either interval or ratio scaled
When would you use chi-square?
non parametric
nominal data
when would you use the Mann-Whitney U-tests?
non-parametric
ordinal
only two groups
all participants are completely randomly chosen
non matched groups
groups are independent
when would you use the Wilcoxon Matched Pairs Signed Rank Test?
Non parametric
Ordinal
Two groups
Matched groups
Two groups are dependent
Within-Subject Design
An experiment where the same group of participants receives more than one treatment of independent variables sequentially.
Example: if you want to see the effect of eating specific foods on memory, the two treatments might be chicken and beef. The group of participants would eat chicken then take a memory test. Later the same group of participants would eat beef and take a memory test.
You would then compare the memory scores to answer the question of what treatment aids memory the most.
