Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
53 Cards in this Set
- Front
- Back
There are more criterion for ____ and _____ people than adults because those two groups are more at risk.
|
children and elderly
|
|
Measurement –
|
Collecting information on which decision will be based
|
|
Evaluation –
|
Using measurement to make decisions
|
|
2 Types of standards –
|
Norm-referenced standards and Criterion-referenced standard
|
|
Norm-referenced standards -
|
Obtained by testing many people in a group and determining what their values are (the norms). Know the group the norm refers to.
|
|
Criterion-referenced standards –
|
predetermined established level to be compared against.
|
|
(Criterion-referenced) You can either:
|
1-Meet or exceed the criterion (sometimes called “passing”) or 2-Fall below the criterion (sometimes called “failing”) Not everybody agrees with the decision.
|
|
Frequency distribution -
|
what is the frequency that a score (or range of scores) is observed
|
|
Statistics
|
Mathematical methods to analyze test scores. Allow us to describe groups of data and the relationship between them
|
|
Histogram:
|
A number of small ranges (bins) of observations on the x-axis (abscissa). The number of observations (frequency) on the y-axis
|
|
Standard shape is a
|
bell
|
|
“Positive skews”
|
refers to a long tail on the positive side.
|
|
“Camel hump/bimodal”
|
refers to two peaks that data is clustered around.
|
|
“Mesokurtic”
|
means a standard bell width.
|
|
“Leptokurtik”
|
means skinny bell curve.
|
|
“Platykurtic”
|
means flat, fat shape (plat, fat, flat)
|
|
Mode –
|
Represents the common numbers
|
|
Median –
|
represents the middle score. Or “50th” percentile. Good for dealing with skewed data or with extreme outliers
|
|
Mean -
|
Average. “x bar”
|
|
Range –
|
difference between the highest score and the lowest score
|
|
~68.3% of all values fall within
|
+/- 1 standard deviations from the mean
|
|
~95.4% of all values will fall within
|
+/-2 standard deviations from the mean
|
|
~99.7% of all values fall within
|
+/-3 standard deviations from the mean
|
|
s=
|
√ ∑(X-Xbar)2 / (n-1)
|
|
__ axis is the independent variable while the __ axis represents is the dependent variable (what we’re often interested in)
|
X...Y
|
|
Correlation Techniques –
|
a mathematical measure of the degree of relationship between two measures (x & y).
|
|
r -
|
Pearson product correlation
A line is fit through the data which sits in a position that minimizes the total deviation, in the x & y direction, from the line to all the data points = line of best fit, = regression line. Range is from +1 to -1. Value indicates direction of relationship, not whether or not is a good relationship. Higher r values allow for prediction of one score. If r=0 there is no relationship |
|
s=
|
√ ∑(X-Xbar)2 / (n-1)
|
|
Positive Relationship –
|
A person with a high measure on one variable tends to have a high measure on the other variabe
|
|
__ axis is the independent variable while the __ axis represents is the dependent variable (what we’re often interested in)
|
X...Y
|
|
Correlation Techniques –
|
a mathematical measure of the degree of relationship between two measures (x & y).
|
|
r -
|
Pearson product correlation
A line is fit through the data which sits in a position that minimizes the total deviation, in the x & y direction, from the line to all the data points = line of best fit, = regression line. Range is from +1 to -1. Value indicates direction of relationship, not whether or not is a good relationship. Higher r values allow for prediction of one score. If r=0 there is no relationship |
|
s=
|
√ ∑(X-Xbar)2 / (n-1)
|
|
Positive Relationship –
|
A person with a high measure on one variable tends to have a high measure on the other variable
|
|
__ axis is the independent variable while the __ axis represents is the dependent variable (what we’re often interested in)
|
X...Y
|
|
Correlation Techniques –
|
a mathematical measure of the degree of relationship between two measures (x & y).
|
|
r -
|
Pearson product correlation
A line is fit through the data which sits in a position that minimizes the total deviation, in the x & y direction, from the line to all the data points = line of best fit, = regression line. Range is from +1 to -1. Value indicates direction of relationship, not whether or not is a good relationship. Higher r values allow for prediction of one score. If r=0 there is no relationship |
|
Positive Relationship –
|
A person with a high measure on one variable tends to have a high measure on the other variable
|
|
Negative Relationship –
|
A person with a high measure on one variable tends to have a low measure on the other variable
|
|
No relationships -
|
a person with a high measure on one variable may have any (i.e. high or low) measure on the other variable
|
|
Correlation Technique –
|
mathematical measure of the degree of relationship between two measures (x & y)
|
|
However, remember correlation does not imply _______
|
causation
|
|
Interpretation of strength of r:
± 0.80 - 1.00 ± 0.60 - 0.79 ± 0.40 - 0.59 ± 0.20 - 0.39 ± 0.00 - 0.19 |
- Very Strong relationship
- Strong relationship - Moderate relationship - Weak relationship - No relationship |
|
How do you decided if a test is a good one? 3 key elements:
|
1. Reliability
2. Objectivity 3. Validity |
|
1. Reliability
|
- Consistent measures (NOT found in the skinfold test). Goal is <5% error
|
|
2. Objectivity
|
- Removes subjectivity
- Two different individuals should be able to arrive at the same score/number - Clearly defined scoring system. - Ways to improve objectivity: remove subjectivity through automation or creating a set of criteria |
|
3. Validity
|
...does the test measure what it’s supposed to measure?
- Logical validity - subjective (expert) decision that the test measures what it claims. –You have a valid method to measure something. - Concurrent validity: see if the measures made by a new test correlates well with the gold standards. - Bioelectrical Impedance Analysis had a strong correlation, but was overestimating. |
|
Factors affecting the reliability of a test:
|
- A test should discriminate throughout the total range of ability: No perfect scores and no zeroes. (Example: When doing a chest press test, make sure that everyone can lift and empty bar and that you have more weight then needed)
- Test only a single attribute (a pull-up is influenced by arm muscles, grip strength, body weight, etc) - Ensure the test is appropriate for the population being tested. (A balance to test older people would result in a perfect score for a college students) - Enjoyablitiy – people wont do it well (or at all) if they don’t enjoy it! |
|
3 Parts of Test Administration
|
1) Practice and warm up must be consistent
2) Directions to give the subject: 3) Scoring |
|
1) Practice and warm up must be consistent
|
-Ensure that all participants have the same amount of practice and the same warm up (beware of the training effect,
- Not just the subject, the proctor needs to practice the test before measuring someone. |
|
2) Directions to give the subject:
|
- Exact directions
- Write out instructions and read them to the subject - Do they get hints to improve performance? - Don’t give them the instructions to read, they may think they don’t need to instructions and so won’t read them. Potentially, they might also misinterpret it. |
|
3) Scoring
|
- Decide what is a successful completion.
- How is an incorrect performance handled? - Score sheet: focus in on necessary data |
|
Before starting data collection on real subject, practice on a “_____” to work out all of the details and the “kinks”
|
friendly
|