Chapter 5 Flashcards
Inferential Statistics
Inferential statistics
- relating the characteristics of a small group (sample) to those of a larger group (population)
- much human performance research is conducted using inferential statistics
- inferential statistics is an extension of the correlational examples
Scientific method
- uses inferential statistics for obtaining knowledge
- requires both the development of a scientific hypothesis and an inferential statistical test of that hypothesis versus another competing hypothesis
Hypothesis
a statement of a presumed relation between at least two variables in a population
Population
the entire group of people or observations in a question (ex. college seniors)
Parameter -population
a measure of interest in the population
Sample - statistics
a study hypotheses about a population by using a subgroup of the population
Ex. there were 200 fifth grades (population), the teacher randomly selected 50 students (sample)
Statistic
the measure of the variable of interest in the sample
Hypotheses are the tools that allow research question to be explored
true
Hypothesis types:
- Research hypothesis
- Null hypothesis
- Alternative hypothesis
Research hypothesis - statement what we expect to happen
what the researcher actually believes will occur
Ex. there will be differences in oxygen uptake based on the type of aerobic training one uses
- researcher can investigate these hypotheses with a t-test or analysis of variance (ANOVA)
Null hypothesis (H0) - statement of equality
a statement that there is no relation (association, relationship, or difference) between variables
population paramenter 1 = population parameter 2
Ex. the mean oxygen uptake is not different for training groups that use different training methods
- hypotheses that you will actually test (and hope to discredit) using the techniques of inferential statistics
Alternative hypothesis (H1)
a statement that there is a relation (association, relation, or difference), typically the converse of H0
population parameter 1 doesn’t= population parameter
Ex. the population mean for group 1 doesn’t equal to the population mean for group 2
- remember that you actually obtain data on sample only and then infer your results to the population
Significance/ alpha level
- probability value (which the results are considered to be statistically significant)
- it allows you to test the probability of the actual occurrence of your result
- the alpha level is set at 0.05 or 0.01 (5% or 1%)
Type 1 error (alpha level)
when the researcher reaches an incorrect conclusion (say there is a relationship or difference when in fact there is not)
- this error sets at 0.05 or 0.01 to make the probability of a type 1 error extremely small
Type 2 error (beta level)
concluding that there is not relation between the variables in the population when in fact there truly is
Chi-square test (X^2)
used to examine associations in nominal data
T-test
used to examine a difference in a continuous (interval/ratio) dependent variable between two and only two groups
ANOVA (analysis of variance)
used to examine differences in a continuous (interval/ratio) dependent variable among two or more groups
Dependent variable
- the criterion variable
- its existence is the reason you are conducting the research study
Ex. strength
Independent variable
- exists solely to determine if it is related to (or influences) the dependent variable
Ex. method of training
if the null is true, the difference between the two means would be zero
true, because there is NO relation
if the probability level is less then 0.05 (ex. 0.01 or 1% of likelihood of type 1 error happening by chance) WE WILL REGECT NULL HYPOTHESIZES
true
if the probability level is over then 0.05 (ex. 0.15 or 15% of likelihood of type 1 error happening by chance) WE WILL ACCEPT NULL HYPOTHESIZES
true
An aerobics instructor is teaching two classes, one in aerobic dance and one in circuit weight training. The instructor wants to know if the proportion of males to females is the same in each class. The correct statistical technique to use is
chi square
Which of the following statistical tests should be used for the following data table of test scores?
Males 85 88 87 81 80 90 91 86 88
Females 90 91 88 89 87 86 90 99 88
t statistic for independent samples
In a one-way ANOVA analysis, which of the following F test statistics is more likely to reject the null hypothesis?
15
Which hypothesis normally is most closely related to the researcher’s hypothesis?
alternative
We make a decision based on probability
true
The normal cures for Null hypotheses will be the same or equal
true
The normal curves for Alternative hypotheses will be different
true
