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;
34 Cards in this Set
- Front
- Back
Central Limit Theorem
|
Data influenced by random effects are approximately normally distributed.
|
|
Variance
|
Indicates the amount of variability in a set of data. Averages of the squared deviations from the mean.
|
|
Standard Deviation
|
Square root of variance. Average amount of dispersion of the scores in a distribution. Small SD = scores not very spread out.
|
|
Descriptive Statistics
|
Provide measure of central tendency. (E.g. mean, median, range, variance, standard deviation.) Analysis of relationships among variables.
|
|
Correlation Coefficient
|
The larger the number the stronger the relationship. + and - indicate the direction of the relationship. Correlation does not equal causation!
|
|
p-value
|
A probability statement which reflects that probability that the effect you tested for occurred as a result of chance.
|
|
Three elements of the product-moment correlation write-up
|
1) A statement regarding the significance and direction (ie. positive/negative) of the relationship.
2) A report of the actual correlation coefficient (r) together with the degrees of freedom (n-1) and the significance level. 3) A description of the nature of the relationship for any significant correlations. |
|
Example correlations write up.
n=15 Mean pure-tone average and word recognition scores: r=-0.94, p<.05 Age and pure-tone average: r=0.09, p>.05 Age and WRS: r=0.10, p>.05 |
There is a significant negative correlation between mean pure-tone average and word recognition scores (presented at 50 dB HL) (r(14) = -0.94, p<.05). Higher pure-tone averages were associated with lower thresholds. Correlations between age and pure-tone average (r(14) = 0.09, p>.05) and between age and word recognition scores (r(14) = 0.10, p>.05) were not significant.
|
|
If you read a question with the word relationship or association you know you need to do which test?
|
Correlation
|
|
Which test would you use: Is there a difference between scores for Groups 1 and 2?
|
Independent t-test
|
|
Which test would you use: is there a difference between pre- and post-test scores for a single group of subjects?
|
Dependent or "paired" t-test
|
|
Independent t-test
|
To determine whether there is a difference between means of two groups on a single condition or attribute.
|
|
Dependent (paired) t-test
|
To determine whether there is a difference between means of a single group on two levels of a dependent variable.
|
|
Which test would you use: are the nine lists of the Modified Simultaneous Sentence Test equivalent in terms of difficulty?
|
ANOVA
|
|
Which test would you use: do people with cochlear dead regions benefit more from digital noise reduction than people without cochlear dead regions (each group tested with DNR and without DNR)?
|
ANOVA
|
|
ANOVA
|
Can be used to determine if there are differences between one or more groups on one or more independent variables with 2 or more levels.
|
|
Which test would you use: differences in thresholds at different frequencies for men and women.
|
ANOVA
|
|
Which test would you use: difference between recognition scores under three different listening conditions for a single group of subjects.
|
ANOVA
|
|
Which test would you use: can PTA predict speech recognition?
|
Correlation
|
|
Which test would you use: is mean hearing loss (PTA) associated with either of the speech recognition measures?
|
Correlation
|
|
A significance level of p<.01 for a correlation coefficient means that
|
There is less than a 1/100 chance that the relationship observed occurred as a result of chance.
|
|
What test would you use: do teenagers with a history of middle ear problems have higher pure-tone thresholds (dB HL) than children without middle ear problems?
|
Independent t-test
|
|
What test would you use: in a group of overweight adult diabetics, does blood glucose decrease after a six-week exercise program (compared to pre-test levels)?
|
Dependent "paired" t-test
|
|
What test would you use: do teenagers with a history of middle ear problems have higher thresholds at 500, 1000, and 2000 Hz (dB HL) than children without middle ear problems?
|
ANOVA
|
|
Type I error
|
False positive. The null hypothesis is true but is incorrectly rejected. Conclude that the observations did not occur by chance, but they really did. The level of significance (p-level)!
|
|
Type II error
|
False negative. Null hypothesis is false, but is incorrectly accepted. Conclude there is no difference, but there really is. Often happens with under-powered designs.
|
|
The probability of making a Type ___ error is the level of significance (p-level).
|
Type I
|
|
You concluded that there is no difference between WRS for people with mild hearing loss and people with profound hearing loss. There really is a difference. This is called a _____ error.
|
Type II
|
|
You concluded that the correlation between height and IQ did not occur by chance. IT really did occur by chance. This is called a _____ error.
|
Type I
|
|
What are you asking in hypothesis testing?
|
"Could these observations really have occurred by chance?"
|
|
Four steps of hypothesis testing
|
1) Formulation of hypotheses: null hypothesis and alternate hypothesis.
2) Test statistic: identify a statistic that will assess the evidence against the null hypothesis. 3) Select p-value: basis for rejecting null. 4) Compare the measured p-value to the selected significance level (alpha). |
|
Null Hypothesis
|
These observations occurred by chance i.e. there is no real difference between groups/conditions or relationship among variables. Generally trying to prove this wrong.
|
|
Alternate Hypothesis
|
There is a real effect i.e. the observations are the results of this real effect plus chance variation.
|
|
f value
|
Systematic error/variance error
|