Data Analysis in Quantitative Research: Understanding Statistics Flashcards
What are the two ways that statistics are used?
- Descriptive (describes the sample)
2. Inferential (outcome variables)
This variable is ________ (discrete or continuous)
Takes on a finite range of values (ex: # of children in household)
Discrete
Is this a discrete or continuous variable?
Also known as categorical variable
Discrete
Is this a discrete or continuous variable?
Used with nominal or ordinal scales
Discrete
Is this a discrete or continuous variable?
Takes on infinite range of values along specified continuum (Ex:Age)
Continuous
Is this a discrete or continuous variable?
Used with interval and ratio scales
Continuous
T/F
Continuous variables and discrete variables can be converted into each other
False. Continuous variables can be turned into discrete variables, but not vice versa
Nominal, ordinal, interval, or ratio?
Lowest level of measurement
Nominal
Nominal, ordinal, interval, or ratio?
Specify rank ordering on variable and assume equivalent distance
Interval
Nominal, ordinal, interval, or ratio?
Mutually exclusive
Nominal
Nominal, ordinal, interval, or ratio?
Attributes are ordered
Ordinal
Nominal, ordinal, interval, or ratio?
Highest level of measurement
Ratio
Nominal, ordinal, interval, or ratio?
Addresses ordering, intervals and absolute magnitude
Ratio
Nominal, ordinal, interval, or ratio?
Does not tell measureable differences between levelsf
Ordinal
Nominal, ordinal, interval, or ratio?
No real absolute magnitude
Interval
Nominal, ordinal, interval, or ratio?
0 means a true absense
Ratio
Parametric vs non parametric
Distribution free
non parametric
Parametric vs non parametric
Considered less robust
non-parametric
Parametric vs non parametric
Easier interpretation
non-parametric
Parametric vs non parametric
unable to interpret multivariates
non-parametric
Parametric vs non parametric
normal distribution
parametric
Parametric vs non parametric
Considered more powerful
parametric
Parametric vs non parametric
Can study multiple variables
parametric
Univariate Bivariate multivariable or multivariate?
One DV, One IV
Univariate
Univariate Bivariate multivariable or multivariate?
multiple DVs
multivariate
Univariate Bivariate multivariable or multivariate?
One DV, multiple IVs
Multi variable
Univariate Bivariate multivariable or multivariate?
One DV, two IVs, or DVs and One IV
Bivariate
Name four nonparametric statistics
Chi-square
Mann-Whitney
Kruskal-Wallis
Friedman
All of the non parametric tests are between-subjects except_______
Friedman
Between-Subjects
non-parametric test
compares an actual number or frequency of a group to the expected number or frequency
nominal data
Between group design
Chi-Squre
This nonparametric test compares two groups. It is between group design and uses ordinal data
Mann-Whitney
This research question indicates which statistical test?
“Is there a difference in breast self examination based on education level?”
Mann-Whitney
This non-parametric test compares more than two groups. It is a between group design and uses ordinal level data
Kruskal-Wallis
This research question indicates which statistical test?
Is there a difference in the medians for change in number of days of having cold symptoms among those who take placebo pill, low vitamin C, and high vitamin C?
Kruskal-Wallis
between groups, ordinal data, more than 2 groups
Nonparametric test
Within groups design
1 group measured on 3+ occasions
Ordinal Level data
Friedman
Which nonparametric test is for nominal data?
Chi-Square
Which nonparametric tests are for ordinal data
Mann-Whitney
Kruskal-Wallis
Friedman
Which research test does this indicate?
Is there a statistically significant difference in perceived sense of well-being survey scores for individuals with clinical depression move from city to rural areas (collected prior to move, 3 months post move, and 6 months post move)
Friedman
What are the two types of t-tests?
- Paired (within groups)
2. Unpaired (between group)
Which research test is indicated?
Is there a difference between treatment and control group in relation to age?
Unpaired t test (independent t-tests)
Why?
continuous data (age)
Between subjects
Which test is indicated?
Is there a difference in mean score of knowledge of breast examination from pre-test to post test?
Paired t-test (dependent t test)
Why?
Continuous data (mean score)
Within subject with pre and post test
When is an ANOVA used?
When comparing three or more group means
What is the statistic computed by ANOVA tests?
F-ratio
Which ANOVA test?
Examines difference between group means of 3+ groups with only one test 1 IV (categorical) 1 DV(continuous) 3 categories of IV
One-way ANOVA
Which test is indicated?
WHat is the effect of therapist intervention, compared to nicotine patch and no treatment on number of cigarettes smoked daily?
One-way ANOVA
Why?
1 categorical IV (Type of therapy)
1 Continuous DV (Number of cigs smoked)
Which ANOVA?
Examines difference between 2 IVs (with 2+ categories each), and one DV(continuous)
Compares each IV with each DV
2 way ANOVA
factorial ANOVA or multifactoral ANOVA
Which test is indicated?
Are counseling, compared to nicotine patch and no treatment equally effective on number of cigs smoked daily for men and women?
2 way ANOVA
Which ANOVA?
measurements taken at more than 2 time points
3+ measures for DV
DV is continuous
Repeated measures ANOVA
Which test is indicated?
What is the difference in depression level of clients recieving mental health services by participating in group therapy sessions at pre test, 6 months and 1 year?
Repeated measures ANOVA
Whch test?
Purpose is to equalize differences between groups by controlling for potentially confounding variables (covariate)
ANCOVA
Which test is indicated?
What is the effect of training on the running speed of adolescents who do or do not work out regularly?
ANCOVA
why? Removes the gender factor IV- training DV-running speed Covariate-gender
Which test?
When 2+ DVs that correlate with each other are compared to 1+ IV
MANOVA
Which test is indicated?
What is the effect of different interventions (desensitization, relaxation therapy, wait-list control) on several types of anxiety (measures iwth 4 subscales on an anxiety meaure)?
MANOVA
why?
DV: 4 subscales of the same test
IV: desensitize, relax, control
Which test?
When 2+ DVs that correlate with each other are compared to 1+ IV
Controls for confounding variables (covariates)
MANCOVA
Which test is indicated?
What is the effect of different interventions (sensitive, relax, cotnrol) on several types of anxiety (measured 4 subscales of anxiety measure) after controlling for age and marital status?
MANCOVA
why?
DV: 4 subscales of the same test
IV: desensitize, relax, control
Covariate: marital status, age
Which test?
A statistical procedure designed to test the relationship or prediction between 2+ IVs and 1 DV
Multiple Regression
Overall fit (R)
Multiple linear Regression
_____ % of mean scores will be within 1 Standard deviation of the mean?
68%
_____% of mean scores will be within 2 standard deviations of the mean?
95%
___% of mean scores will fall between 3 standard deviations?
99%
The mode of the slope is to the right, but the tail and mean of the slope points to the left.
Negative skew or positive skew?
negative skew
The mode of the slope is to the left, but the tail and the mean of the slope is to the right
Positive skew
What is the purpose of bivariate descriptive statistics?
To describe relationships between two variables
What are some examples of bivariate descriptive statistics?
- Correlation
- Absolute risk
- relative risk
- relative risk reduction
- odds ratio
T/F Correlational Pierson’s product use variables on nominal and ordinal scales
False, interval and Ratio
Which correlational test measures ordinal variables?
Spearman’s Rho
The proportion of people who experience an undesirable outcome in intervention
Absolute Risk
Estimated proportion of an adverse outcome that persists when people are exposed to the intervention
Relative Risk
The proportion of people with the adverse outcome relative to those without
Odd Ratio
T/F Rejecting the null hypothesis means the alternative hypothesis is true
False. It might be true
A type one error means?
False positive:
The researcher rejects the null hypothesis when he should not
What is a type two error?
False negative
The researcher fails to reject the null when he should have
How do you control for a type one error?
Descrease the significance level
How do you control for a type 2 error?
Power Analysis