chapter 2/week 2 Flashcards
behavioral variability and research
variability
the degree to which scores in a set of data differ or vary from one another
Central to the research process
variability (list) - no variability no story
- Psychology and other behavioral sciences involve the study of behavioral variability
- Ex 1 - Dating
- Research questions in all behavior sciences are questions about behavioral variability
- Ex - Differences in dating behavior among generations?
- Studies show be designed in a way that best allows the researcher to answer questions about behavioral variability
- Ex - sample 80 people around 20, sample 80 people around 20; survey, t test, compare
- The measurement of behavior involves the assessment of behavioral variability
- Ex - how are we going to measure dating; commitment? Frequency of dating?
- Statistical analyses are used to describe the amount for the observed variability in behavior
Variability and Research steps
- focus on variability
- questions about variability
- design for variability
- measuring variability
- statistical analyses for variability
Variability and Research - 1. Focus on Variability
What is the source of variation?
Is it variation/change
- Across situations
- Among individuals
- Over time
Variability and Research - 2. Questions about Variability
Identify what your research questions is about
- What varies
- To what extent can we predict variability in a certain behavior
- Why does it vary
Variability and Research - 3. Design for Variability
Research should be designed in a manner that best allows the researcher to answer questions about behavioral variability
Think of your questions
- How does it vary
- Observation, correlational
- Under what circumstances does it vary - Quasi-experimental - How can we control variability - Experimental
Variability and Research - 4. Measuring Variability
The measurement of behavior involves the assessment of behavioral variability
We measure behavior through operationalization of concepts
We assign numbers to behaviors such that the variability in the numbers reflects the variability in the behavior
No variability, no assessment
Variability and Research - 5. Statistical Analyses for Variability
Statistical analyses are used to describe and account for the observed variability in the behavioral data
We use data analysis to answer questions about the variability in our data:
How much variability is there
What is it related to
What caused it
An Illustration of Hypothesis Testing
General Hypothesis example – There is an association between one’s stress level and having to deal with contradictory views at once
Null Hypothesis example – there will be no association between having to deal with contradictory views and one’s general comfort level
Two Types of Statistics
descriptive statistics
inferential statistics
descriptive statistics
used to summarize and describe the behavior of participants in a study
inferential statistics
used to draw conclusions about the reliability and generalizability of one’s findings
How to capture variability in behavior using statistics?
range
variance
standard deviation
range
difference between the largest and smallest scores (observed participant behavior) in a distribution
variance
a statistic that takes into account all the scores when calculating the variability
a statistic used to indicate the amount of variability in participants’ responses
standard deviation
the square root of the variance; generally easier to interpret
how to calculate standard deviation
find the mean
subtract mean from each score (deviation)
Then square each of the values (the differences)
Calculate the total sum of the squares [add the squared values together]
Calculate the variance – divide the sum by the number of values present minus 1
take square root of the result
why is Variance A Better Alternative ?
range doesn’t describe the distribution
We express the variability of the data using all the scores, not just the highest and lowest ones, with a statistic called the variance
mean
the sum of a set of scores divided by the number of scores
Σyi / n
variance and the mean (interpretation)
We assess the variability in a set of data by seeing how much the scores vary around the mean
If the scores are tightly clustered around the mean, then the variance of the data will be small
If the scores are more spread out from the mean, then the variance will be larger
how to calculate variance
Calculating the mean of the data
Subtracting the mean from each score (deviation)
Squaring these differences or deviation scores
Summing these squared deviation scores
Dividing by the number of the scores minus 1
statistical notion of variance
s² = Σ (y_{i} - ȳ)²(n-1)
deviation
score:
- how the scores vary around the mean
- how each score differs from the mean
positive deviation score
indicates that the participant’s response fell above the mean
negative deviation score
indicates that the participant’s response fell below the mean
normal distribution
normal curve
split:
34.1%, 13.6%, 2.1%, 0.1%
-3σ, -2σ, -1σ, mean/mu, 1σ, 2σ, 3σ
total variance
the total sum of squares divided by the number of scores minus one
Total variance = systematic variance + error variance
systematic variance
the portion of the total variability in participants’ scores that is an orderly, predictable fashion to the variables the researcher is investigating
This phrase refers to the part of the variation in participants’ scores (or outcomes) that can be explained or predicted by the specific factors or variables the researcher is studying.
error variance
the portion of the total variance in participants’ scores that is unrelated to the variables under investigation in the study; variance remains unaccounted for
AKA - measurement error, experimental error
–> Can mask or obscure the effects of the variables in which researchers are primarily interested in
Distinguishing Systematic from Error Variance
Researchers use statistical analyses to partition the total variance of their data into systematic and error components
The more error variance in the data, the more difficult it is to determine whether the variables of interest are related to variability in behavior
Researchers try to minimize error variance as much as possible in order to detect the systematic variance in the data
effect size
indicates the proportion of the total variance that is systematic variance
A measure of strength of association - the strength of the relation between two variables
Because the effect size is a proportion, it is easy to compare effect sizes across many different studies with different research strategies
Assessing the Strength of Relations
If effect size = .00, then none of the variance in participants’ responses is systematic
If effect size = 1.00, then all the variability in the data can be attributed to the variables under study
The larger the proportion, the stronger the relationship between variables
There are several indicators of Effect Size:
Correlational effect sizes: r-squared
Group difference effect size: Cohen’s d
Association among categorical variables: odd-ratio
meta-analysis
a procedure used to examine every study that has been conducted on a particular topic to assess the relationship between whatever variables are the focus of the analysis
Effect sizes allow us to compare across studies: meta-analysis
By looking at effect sizes across many studies, a general estimate is calculated to reflect the strength of the relationship between the variables
Research is a quest for Systematic Variance
All researchers are trying to account for (or explain) the behavioral variability they observe
If there is no variability, there is no systematic variance and hence no story to be told
