Chapter 16 Pwpt Flashcards
Measurement is?
Assignment of a number
The nature of the variable or event determines the _________ ______ _________?
Level of measurement
The level of measurement determines the ______ __ _________
Type of statistics
The four levels of measurement are?
Nominal
Ordinal
Interval
Ration
Nominal is?
Classification of variables
Lowest level
Least amount of manipulation
Dichotnomous and categorical
Dichotonomous and categorical mean?
Dichotonomous= only 2 true variables Categorical= more than 2
Ordinal? And examples?
Rankings of high or low
Frequencies, percentages, medians, percentile, rank order
Interval? And examples?
Rankings with equal intervals between allows more manipulation
Ex. Likert, quality of life, depression, functional status
Ratio? And examples?
Rankings of events or variables on scales with equal intervals and absolute zeros
Highest level of measurement
All mathematical procedures can be preformed on ratio measurements
Ex. Height, weight, pulse, bp
Descriptive statistics?
Describe and summarize
Measures central tendency (mean, median, mode)
Measures variability (range, standard deviation)
Correlation techniques (scatter plots)
Inferential stats?
Predict and generalize
Analyze, test hypotheses, answer research questions
Used to draw conclusions that extend beyond immediate data of the study
Descriptive statistics: frequency distribution
Basic way to organize data
Counts the number of times each event occurs
Measures of central tendency?
Single number describes the middle of the group
AKA summary statistics- appropriate measure depends on the level of measurement
Mode, median, mean?
Mode= most frequent
Median=middle
Mean=average
Modality means?
Number of modes contained in a distribution
When all scores in a distribution are different, it is possible to
Have no mode
Normal distribution?
Symmetric distribution
Sometimes bell shaped
Non symmetrical distribution?
Peak of curve off center
Positive or negative skews
Descriptive stats, measures of validity?
Spread of data Homogeneity? Range Semiquartile range Percentile Standard deviation
Inferential statistics
Allow the testing of hypotheses using data obtained from probability and nonprobability samples
Two hypotheses are tested (research/scientific and null)
Probability?
What are the chances of obtaining the same result from a study that can be carried out many times under identical conditions?
Repeated trials allows probability to test hypotheses
Sampling error
Types of errors
Type 1 error info
More serious
Can cause a type 2 error
Type 2 error
Occurs when sample is too small
May limit opportunity to measure the treatment effect (true difference between two groups)
A larger sample improves the ability to detect the treatment effect
Inferential statistics: level of significance?
Alpha level
Probability of making a type 1 error
Probability of rejecting null hypotheses
Minimum level of significance acceptable is: 0.05
Inferential statistics; statistical significance vs. clinical significance
If statistically significant, unlikely to have occurred by chance
Does not necessarily imply clinical relevance
Tests of significance: nonparametric
Less powerful and flexible
Used with nominal and ordinal
No estimation of population parameters
Distribution of data is skewed
Tests of significance: parametric
More powerful and flexible
Used with interval or ratio variables
Estimate of at least one population parameter
Variable is normally distributed
Tests of difference
Most commonly used in experimental and quasi experimental designs
t-test: statistically tests whether two groups means are different
Analysis of variance, ANOVA?
More than two groups or measurements taken more than once
Analysis of covariance (ANCOVA)
Used to measure difference among group means, but it also uses a statistical technique to equate groups under study on an important variable
Multiple analysis of variance (MANOVA)?
Used to determine the differences in group means, but it used when there is more than one dependent variable
Chi square is used when?
Used when data are at the nominal level and the researcher wants to determine whether groups are different
Nonparametric
If samples are small and expected frequencies are less than 6 inches each cell what probability test is used?
Fishers exact probability
Tests of relationships do what?
Explore the association or correlation between two or more variables
Usually associated with nonexperimental designs that provide level IV evidence
Pearson product moment correlation is associated with what level of measurement?
Parametric level of measurement,
Interval or ratio
Phi coefficient is used?
To express relationships when two variables being tested have only two levels
Nonparametric: nominal
Point-biserial correlation is used?
To determine relationship between a nominal variable and an interval variable
Nonparametric
Nominal
Spearman Rho is used?
To determine degree of association between two sets of ranks as is Kendall’s tau
Nonparametric
Ordinal
Effect sizes? Small, medium and large?
Small: r=0.1
Medium: r=0.3
Large: r=0.5
Advanced statistics: multivariate statistics are?
Multiple regression
Factor analysis
Path analysis
Structural equation modeling
Appraising the evidence, descriptive and inferential stats?
What descriptive statistics are reported?
What level of measurement is used to measure each of the major variables?
Is the sample size large enough to prevent one extreme score from affecting the summary statistics Does the hypothesis indicate that the researcher is interested in testing for differences between groups or in testing for relationships? What is the level of significance?
Does the level of measurement permit the use of parametric statistics?
Is the size of the sample large enough to permit the use of parametric statistic?
Are the results for each of the hypothesis presented clearly and appropriately?
If tables and graphs are used, do they agree with the text and extend it, or do they merely repeat it?
Are the results understandable?
Is a distinction made between clinical significance and statistical significance? How is it made?
Science and research…..? One studies findings are rarely sufficient to support a major practice change
Prove nothing.
Identify the level of measurement for the variable of age?
Interval
To minimize the risk of a type 2 sample error, the researcher should do what?
Increase the sample size
Which test is used to test for differences between means?
t test
Which statement regarding statistical hypothesis testing and errors is true?
Type 1 error occurs when the researcher rejects a null hypothesis that is actually true
