Statistics Flashcards
Order of Research Hierarchy
• Systematic Review &
Meta-Analysis
• Randomized Controlled
Trials
- Cohort Study
- Case study
- Expert opinions
What is an Independent Variable?
Independent variable (Intervention)
• A condition, intervention, or
characteristic that will
predict or cause a given outcome.
What is a Dependent Variable?
Dependent variable (outcome)
• The response or the effect
that presumed to vary
depending on the
independent variable
What is a Null Hypothesis?
- Null Hypothesis (Ho)
- Why do we need this
- Establishes a possible explanation for
the results of the experiment
- What is a null hypothesis
- No difference
- Hypothesis conclusion
- If I “Do Not Reject” the null hypothesis,
I am saying there is no difference
• If I “Reject” the null hypothesis,
I am saying that there is a difference
What is Significance Value (Alpha-value)?
We set this at the beginning of our
research.
The risk that we are willing to take that
results are due to chance.
- Example
- α = 0.01 = 1%
- α = 0.02 = 2%
- α = 0.05 = 5%
- α = 0.1 = 10%
What is the Probability Value (P-value)?
the probability of finding a significant
change when there actually isn’t one
present.
“The probability of making a mistake”
- Example
- If we set α = 0.05 = 5% and p < 0.05
- There is a REAL change and we
[reject] the Null Hypothesis
Type 1 error?
• Smooth Criminal
Stating that there is a significant
difference when no true difference
exists
- Why is it important?
- Stating that a certain intervention is
superior for improving balance when it
isn’t can delay delay patient progress
- Example
- Incorrectly rejecting the null
hypothesis
What is the Beta Value?
the probability of stating
that there is no real change
when there actually is a real
change present.
- Example
- B= 0.20 = 20%
What is a Type 2 error?
- Hiding In Plain Sight
- Stating that there is not a real change when
there is a real change
- Why is it important
- Stating that a certain intervention does not
yield a significant difference when it does
- Example
- Failing to reject the null hypothesis
What is Power?
• The probability that
we will be able find a
difference when one
truly exists.
- Example,
- 80% is normal
FACTORS THAT INFLUENCE POWER:
Sample Size
• The larger the sample, the
greater the statistical power.
- Variance
- The power is increased as the
variance within a group is
reduced 1 2 3 4
- Significance Criterion
- The power is increased as the
alpha value is increased Alpha Value set to 0.05, 0.1, or higher
- Effect size
- Amount of significant change
observed
What is a dependent (paired) T test?
- Statistical Test
- Used to determine the difference within the same
person or same group of people
• Pre intervention & Post-intervention
What is an independent (unpaired) T-test?
Statistical Test
• Used to determine the difference between two
independent groups of data
One Tailed versus Two-Tailed T-Test
Dependent on your initial
hypothesis
• If you hypothesize that
that the 8-week
quadriceps program will
have a longer single leg
stance time than the
general exercise group
then you run a one tail ttest.
• If you hypothesize that
that there will be a
difference between the
two groups but you don’t
know if the 8-week
quadriceps program will
have a longer or shorter
single leg stance time than
the general exercise group
then you run a two tail ttest.
What is ANOVA used for?
Statistical procedure appropriate for comparison of three or more treatment groups or conditions
What is a Bell Curve Skew?
Levels of measurement
Nominal – categories with no ranking
Ordinal – categories with intrinsic rank
ordering with no magnitude of
difference defined
Interval - magnitude of difference is
defined; but zero is undefined
Ratio – magnitude of difference is
defined; and zero is meaningful
Intraclass Correlation Coefficient
About CORRELATIONS
-1 is negative/inverse relationship (when one goes down, the other goes up)
0 is no correlation
+1 is when one goes up, so does the other, positive correlation
Pearson Coefficient
The Pearson coefficient is a type of correlation coefficient that represents the relationship between two variables that are measured on the same interval or ratio scale. The Pearson coefficient is a measure of the strength of the association between two continuous variables. This is measured similarly to ICC.
Chi-Square
The chi-square statistic compares the size any discrepancies between the expected results and the actual results, given the size of the sample and the number of variables in the relationship. For these tests, degrees of freedom are utilized to determine if a certain null hypothesis can be rejected based on the total number of variables and samples within the experiment
