Descriptive stats Flashcards
What are the 3 research designs?
Deductive / inductive
Quantitative / qualitative
Design quality
What is deductive approach?
Start from a theory that will be tested
Th => Hyp => Obs => confirmation / rejection
What is inductive approach?
Data guides theory
Obs => pattern => Th
What is quantitative approach?
Manipulating data analyzed using mathematically based methods.
What is qualitative approach?
Results obtained are narrative / descriptive.
What is the difference between observation and experimentation?
Obs: measuring variables without intervention of the researcher.
Exp: Measuring manipulating at least 1 variable.
What are the 4 criteria of design quality?
Validity, Reliability, Replicability, Generalizability
What is Validity?
Do I measure what is supposed to be measured?
What is Reliability?
Does this research produce stable results?
What is Replicability?
If someone practices the same study exactly, will he find the same results?
What is Generalizability?
Does this experiment also occurs in the real life?
Is it possible to practice inference?
What are variables / indicators?
People or things from which data is collected.
Statistical units / cases.
What are the different types of variables?
Nominal / categorical (the less informative)
Ordinal
Interval / scale
Ratio (the most informative)
What is a nominal / categorical variable?
categories without order.
What is an ordinal variable?
2 or more categories that can be ordered.
What is an interval variable?
The difference between 2 values of this variable is meaningful.
What is a ratio variable?
Each interval between 2 numbers is meaningful.
This allows to practice operations because there’s an absolute zero.
What are the roles of variables?
Independent Variable (predictor / var predicted)
Dependent Variable
Each role leads to a different sort of analyses.
What is an Independent Variable?
Variable manipulated / observed to observe the effect on the DV. It can be measured.
What is a Dependent Variable?
What is measured.
What are the 2 main fields for statistical analyses?
Sample => descriptive statistics
Population through informations obtained on the sample => inferential statistics
What are the 2 possible inferential statistic methods?
Calculate a difference from 0 => p ratio.
Defining an interval in which the population is likely to be => the confidence interval. (Better)
How are called statistics analyses regarding the number of variable involved?
1: Univariate
2 simultaneously: Bivariate
More than 2 simultaneously: Multivariate.
What is the difference between parametric and non-parametric test?
Parametric tests are passed on the assumption of a normal distribution.
Non-parametric test can be used to test any distribution, even when it’s not normal.
What are the 2 main attentions you have to give to the plan data before you start collecting data?
Sample carefully selected: randomly
Sample size: you must know the test before you gather data to choose the perfect size of sample.
Outcome = ?
Outcome = model + error
=> Error = Outcome - model
What does mean “calculating the parameters”?
Finding the values of the parameters that will minimize the error / deviation / residuals.
What is used to calculate the parameters?
SSE = ⁿΣi=1(Obs-pred)²
It must be the lowest possible. This is the method of ordinary least squarer.
What is the variance formula?
V = 1/N(ⁿΣi=1(obs-mean)²)
