29 Jul 2016

School

Department

Course

Professor

Today

Statistics are tools, not fear-inducing instruments of torture

Some basics of measurements

No previous work in the area, new data, we just want to explore what happens

○

Tends to be more qualitative

○

Exploratory

•

Want to understand and describe the state of a population

○

Descriptive

•

We want to understand cause and effect, it explains something

○

A lot of quantitative data, but can use qualitative by using open ended questions

○

Explanatory

•

Using one piece of information to predict another

○

Predictive

•

Data's purpose is…

Data can be quantitative or qualitative

•

Data's purpose is multiple in nature and it depends on what we are doing with our research question

Statistics

Frequencies, a graphic representation, central tendency, distributions

•

Describing a sample (the average population age)

•

Descriptive

Probability of something occurring , where data standards in relation to normal, t-tests, analyses

•

Inferential

The Language of Statistics and Definitions and Scaling

Saturday, May 28, 2016

8:51 PM

Second Term Page 1

Probability of something occurring , where data standards in relation to normal, t-tests, analyses

of variance, correlations, regression ,chi-square

•

The tools we use ( formulas) to make inferences about a population based on our sample

•

Inference

Experimental statistics, the kind of work in which we want an explanation

•

Casual inference

Observation comes from a sample, which forms our data

•

Stats makes us make our data observation inference to make a conclusion

•

Reasoning from an observation to a conclusion

Population

Everyone in the group of interest

Broad as all adult humans or as narrow as only females between 18-24 that are taking stats

at western

○

Everyone

•

Any information about the population

○

Parameters

•

A population is a complete collection of anything, and a parameter is a numerical characteristic of a

population.

Sample

A sample is a subset of the population; a measurable characteristic of a sample is called a statistic.

Sampling is the process of selecting sample from the population. Random sampling is a method in which

each member of the population has an equal chance of being selected. In stratified random sampling,

the population is divided into relevant groups (strata), and then random samples are taken from each

group.

Generalizations

•

Any information about the sample

○

Statistics

•

The representative group --> the individuals within that population that you are subjecting to your

testing procedure

If it isn't representative you are lacking external validity

○

Must be a valid representation of the population

○

Representative sampling = external validity

•

Random sampling

•

Availability sampling (convenience sampling)

•

It's all about the sample

Experimental Study

You manipulate in order to cause an effect

•

Manipulation study

A variable is anything that may take on different values or amounts.

The independent variable in an experiment is the variable manipulated and controlled by the

experimenter to determine it effect on the dependent variable. In the behavioral sciences, the dependent

variable is the measurement of behavior.

Levels --> minimum of two levels

•

This is the variable you are manipulating

•

Independent variable(s)

Second Term Page 2

This is the variable you are manipulating

•

The variable that is being measure

•

This is where we can see the effect

•

Dependent Variable(s)

The goal is…

To make analytical comparisons

Measurement error (tool, malfunction)

○

Researcher error

○

Participant error - reaction time is slower from being up the night before, etc

○

Error is about variability that we can't account for based on our manipulation

•

Any differences not accounted for by manipulation

○

E.g., intelligence differences, speed differences, sex differences, personality

○

Differences in the dependent variable between groups

•

Statistically this is reduced to…

The smaller the error, the greater the ratio

Non-Experimental Study

Relationships that are predictable

Variables in a world of mediators and moderators

The goal is…

To find associations and predictors

Statistically this is reduced to…

What we're going to learn

The language of statistics

How we approach statistics

We then have the ability to look at a research question and know if you should be using a T-

test or an Anova

○

Which is the rationale for the statistics used

•

By explaining the theory

Second Term Page 3