Research methods 3
Main primary display types
1. frequency table 2. Histogram
Columns in a Frequency Table
1st column- depicts different categories or levels of a variable 2nd column- depicts the frequency of each level
Histogram
A graph of vertical bars representing the frequency distribution of a set of data.
What is a mixed- design ANOVA?
A search design in which at least one variable is between- groups and at least another is withingroups.
descriptive statistics
A set of techniques for summarizing and displaying data. Usually giving information about one variable
Effect size
A statistic which describes the strength of a statistical relationship
Why do you use descriptive statistics
All research uses this technique to give people an idea of what participants looked like within that variable. % female, mean age
The number of paid days off taken by 8 employees at a small local business is compared to the national average of 15 days. You want to figure out whether the employees from the small local business are getting about as many as the national average. The data for those 8 employees are: 10, 11, 8, 14, 13, 12, 12, and 27 days. • Assume the obtained value (i.e., the result) from this test is -0.79, p = .897, and the effect size is d = -0.28. Interpret these results in APA style. That is, do we reject or retain the null hypothesis? Is there a statistically significant difference in number of days of leave between the employees of this small business and the national average? Does this small business need to talk to their employees about how many days they are absent so they can restrict the number of days they take on leave? (Assume alpha is .05).
Although the average number of paid days off at this small business is lower than the national average, it is not statistically significantly lower
What is a two- way ANOVA?
An ANOVA with two IVs
In a factorial design, what are all possible options for assigning participants to the different conditions? Please name them AND provide the definition.
Between-subjects - each person in assigned to one condition Within-subjects - each person is assigned to all conditions Mixed factorial - each person receives one condition of one IV and all conditions of the other IV
independent samples t-test
Compares the means between two groups accounting for their SD and sample size
Frequency table
Displays distribution of a variable, it organizes categorical or numerical data and how often each occurs. (Table form)
What does Cohen's d measure?
Effect size. The magnitude of the relation or difference we are measuring regardless of the sample size.
How can we turn a between groups design to a within groups design?
Get all the participants to be in all the levels of the IV.
Order effects
How a participant's behavior changes when the dependent variable is presented for a second time.
Main effect
In a factorial design when an IV has an influence on the DV
Interaction
In a factorial design when the effect of one or more levels of an IV is strengthened or weakened at one or more levels of the other IV.
Why is an order effect considered a confounding (or third) variable?
It introduces error or noise to our results. That is, it changes our DV in ways that have nothing to do with the IV. Therefore, it makes it harder for us to conclude that changes in the DV were caused by the IV we are interested in because an order effect has introduced one or more IVs to our study that we might not be able to control.
When do we use a single sample t-test?
It is a hypothesis test in which we compare a sample from which we collect data to a population for which we know the mean but not the standard deviation.
What is a Post Hoc test?
It is a statistical procedure frequently carried out after the null hypothesis has been rejected in an analysis of variance (ANOVA).
(M1 - M2)SD pooled
Mean from group 1 - mean from group 2 X average SD of both groups
What are two situations to consider when using correlations.
Non linear relations Restriction of range
A researcher believes that when students receive information about the nutritional value of their food, they will start making healthier food choices. Specifically, she hypothesizes that students are likely to avoid healthy lunches by avoiding vegetables and smuggling in banned junk food unless they receive accompanying nutrition education and hands- on involvement in their meals. To test this, she has students randomly assigned to either participate in the Edible Schoolyard program in which they receive education and hands- on training or to continue with their usual lunch plan. After the study is over, she measures whether their weight changed. What would be our null and alternative hypotheses?
Null hypothesis: Students who participate in the Edible Schoolyard program weigh the same, on average, as students who do not participate Alternate hypothesis: Students who participate in the Edible Schoolyard program have different weights, on average, than students who do not participate
What is the effect size when using correlational analyses?
Pearson's Correlation Coefficient (r) Indicates IF they are different and HOW much they are different
What are two ways to measure variability in data?
Range and Standard Deviation
The number of paid days off taken by 8 employees at a small local business is compared to the national average of 15 days. You want to figure out whether the employees from the small local business are getting about as many as the national average. The data for those 8 employees are: 10, 11, 8, 14, 13, 12, 12, and 27 days. What test will you use to calculate whether this sample is different than the national average? Briefly explain why?
Single sample t-test, one sample compared to one population
Under which research designs might order effects occur?
Sometimes called a practice effect. That is, when the participant has been in more than one level of the IV (within groups design), the participant is also receiving the test for the DV more than once (practice makes for better results. Also, that score may also have been affected by the treatment or activities that were presented in the previous IV level the participant may have experienced.
Strength and Cohen's D
Strength Cohen's D Large ± 0.80 Medium ± 0.50 Small ± 0.20
Strength and pearson's R
Strength. Pearson's r Large. ± 0.50 Medium ± 0.30 Small ± 0.10
How are differences in groups determined
The Means for each group are compared • If the means are statistically different (p <.05) they are considered to be different from each other. The standard deviations (SD) are compared If the SD for each group is not too variable, this will help the groups be statistically different
Variability
The extent to which the scores in a data set tend to vary around the central tendency.
A researcher believes that when students receive information about the nutritional value of their food, they will start making healthier food choices. Specifically, she hypothesizes that students are likely to avoid healthy lunches by avoiding vegetables and smuggling in banned junk food unless they receive accompanying nutrition education and hands- on involvement in their meals. To test this, she has students randomly assigned to either participate in the Edible Schoolyard program in which they receive education and hands- on training or to continue with their usual lunch plan. After the study is over, she measures whether their weight changed. What is the independent variable (s), the level(s) of the independent variable(s), and the dependent variable?
The independent variable is the program, with two levels (Edible Schoolyard, control), and the dependent variable is weight after the program.
The number of paid days off taken by 8 employees at a small local business is compared to the national average of 15 days. You want to figure out whether the employees from the small local business are getting about as many as the national average. The data for those 8 employees are: 10, 11, 8, 14, 13, 12, 12, and 27 days. Write out the null and alternate hypotheses
The null hypothesis is that this small business is the same as national businesses in terms of average number of paid days off The alternative hypothesis is that there is a difference in average number of paid days off for this business compared to other businesses nationwide
When do we use an independent t-test? Why is it called an independent test?
To compare two means for a between- groups design, a situation in which each participant is assigned to only one condition. It is independent because the scores that we used to obtain the two sample means come from different participants. Otherwise put, the sample means come from two different populations.
When do we use a paired- samples t-test
To compare two means for a with-in groups design, a situation in which every participant is in both samples; also called a dependent- samples t-test
Example of distribution
Variable = race levels = Caucasian, African American, hispanic/ latino, pacific islander etc
In null hypothesis significance testing (t-test and F-tests) when the obtained statistic is larger than the alpha threshold (i.e., the critical value), what decision do we make about the null hypothesis and the effects of the IV on the DV?
We reject the null hypothesis, and conclude there is a statistically significant effect of the IV on the DV.
A researcher believes that when students receive information about the nutritional value of their food, they will start making healthier food choices. Specifically, she hypothesizes that students are likely to avoid healthy lunches by avoiding vegetables and smuggling in banned junk food unless they receive accompanying nutrition education and hands- on involvement in their meals. To test this, she has students randomly assigned to either participate in the Edible Schoolyard program in which they receive education and hands- on training or to continue with their usual lunch plan. After the study is over, she measures whether their weight changed. What would happen to our interpretation of the results if this researcher had not randomly assigned her participants to the two groups? Would we be able to assume causality? Why or why not?
We would not be able to assume high causality (there is still some internal validity though) because this would not be a true experiment (it would be a quasi- experiment)
A researcher believes that when students receive information about the nutritional value of their food, they will start making healthier food choices. Specifically, she hypothesizes that students are likely to avoid healthy lunches by avoiding vegetables and smuggling in banned junk food unless they receive accompanying nutrition education and hands- on involvement in their meals. To test this, she has students randomly assigned to either participate in the Edible Schoolyard program in which they receive education and hands- on training or to continue with their usual lunch plan. After the study is over, she measures whether their weight changed. What test would we use to determine whether the Edible Schoolyard program makes a difference? Why?
We would use an independent-samples t test because there are two samples and no student is in both samples.
What is a matched- groups research design?
When we can not have a within groups design we try to match the participants from one group to those in the other group on as much characteristics as possible (e.g., various demographic information, health, intelligence, etc.)
When do we use a one- way between groups ANOVA?
When we have one nominal IV with more than two levels, each sample is composed of different participants, and we have a scale DV.
When do we use a one way within- groups ANOVA?
When we have one nominal IV with more than two levels, each sample is composed of the same participants, and we have a scale DV. This design is also called a one- way repeated measures ANOVA.
Standard Deviation
a measure of variability that describes an average distance of every score from the mean
Factorial design
a study in which there are two or more independent variables, or factors
What does ANOVA stand for?
analysis of variance
central tendency
distribution of the middle- the point where scores of a distribution tend to cluster
What does a p-value of .10 mean? - The answer is not simply "it's not significant"
if the null hypothesis were true (no effect/relationship), results this extreme would occur only 10% of the time.
Three types of central tendency
mean, median, mode
range
the difference between the highest and lowest scores in a distribution
null hypothesis
the hypothesis that there is no significant difference between specified populations, any observed difference being due to sampling or experimental error.
Distribution
the way that scores are dispersed across levels of a variable/ how many people fit into each possible level ( category) of the variable
What is ANOVA?
type of null hypothesis significance test typically used with one or more nominal IV ( with at. least three groups overall) and scale DV
Why should we always calculate an effect size?
• Because the effect size gives us more information about the relations regardless of sample size