Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
37 Cards in this Set
- Front
- Back
Statistics
|
Statistics are used to summarize large data sets and to extract meaning from them
|
|
Population
|
Population: all of the individuals of interest in a study
examples: • all people with bipolar disorder • all incarcerated people with bipoloar disorder • all men with bipolar disorder • all men • all people |
|
A variable
|
is any measure that can take on more than
one value in a set of data (numerical information) |
|
Dependant variables
|
– Proportion of voters favouring the Conservatives
– Number of voters favouring the Conservatives – Average age of voters favouring the Conservatives – Average age of voters |
|
Independent variables
|
– Political party (Conservative, NDP, Liberal, Green)
– Region of Canada (BC, Prairies, Ontario, Québec, Atlantic, North) – Date of poll (April 1, 2, 3, 4…) |
|
Variables can be measured on one of four scales:
NOIR |
– Nominal
– Ordinal – Interval – Ratio |
|
Nominal scale
|
values are arbitrary, no math
possible |
|
Ordinal scale:
|
values have only relative meaning, can only be rank ordered
--Differences between consecutive rankings have no meaning – Can’t meaningfully perform mathematical operations on ordinal data ~ we do not know exactly how much something is more than another item on a scale - differences between consecutive rankings have no set meaning - can’t meaningfully perform mathematical operations on ordinal scales |
|
Interval scale
|
Order is meaningful (just as for ordinal data) but now the differences between values are meaningful and constant
– Interval scales have an arbitrary zero point – Can’t compare magnitude, only direction ~ only scale that allows for negative values i.e.) temperature |
|
Ratio data
|
Retains all of the properties of interval data (order is meaningful, intervals are constant)
but zero point is not arbitrary • Because 0 means “absence of” on this scale, it is meaningful to compute ratios, percentages, etc. • The appropriate statistical test depends on the type of data we have i.e.) weight: when you have zero pounds you have nothing |
|
Statistics
|
Statistics are used to summarize large data sets and to extract meaning from them
|
|
Population
|
Population: all of the individuals of interest in a study
examples: • all people with bipolar disorder • all incarcerated people with bipoloar disorder • all men with bipolar disorder • all men • all people |
|
A variable
|
is any measure that can take on more than
one value in a set of data (numerical information) |
|
Dependant variables
|
– Proportion of voters favouring the Conservatives
– Number of voters favouring the Conservatives – Average age of voters favouring the Conservatives – Average age of voters |
|
Independent variables
|
– Political party (Conservative, NDP, Liberal, Green)
– Region of Canada (BC, Prairies, Ontario, Québec, Atlantic, North) – Date of poll (April 1, 2, 3, 4…) |
|
Variables can be measured on one of four scales:
NOIR |
– Nominal
– Ordinal – Interval – Ratio |
|
Nominal scale
|
values are arbitrary, no math
possible |
|
Ordinal scale:
|
values have only relative meaning, can only be rank ordered
--Differences between consecutive rankings have no meaning – Can’t meaningfully perform mathematical operations on ordinal data ~ we do not know exactly how much something is more than another item on a scale - differences between consecutive rankings have no set meaning - can’t meaningfully perform mathematical operations on ordinal scales |
|
Interval scale
|
Order is meaningful (just as for ordinal data) but now the differences between values are meaningful and constant
– Interval scales have an arbitrary zero point – Can’t compare magnitude, only direction ~ only scale that allows for negative values i.e.) temperature |
|
Ratio data
|
Retains all of the properties of interval data (order is meaningful, intervals are constant)
but zero point is not arbitrary • Because 0 means “absence of” on this scale, it is meaningful to compute ratios, percentages, etc. • The appropriate statistical test depends on the type of data we have i.e.) weight: when you have zero pounds you have nothing |
|
parameters
|
Numbers used to summarize populations of scores are called parameters
– The value of a parameter is always what we’re interested in, and rarely directly measureable Parameter: is a value, usually a numerical value that describes a population. Parameter is usually derived from measurements of the individuals in the population - a characteristic that describes the population i.e.) the average score for a population |
|
statistics
|
Numbers based on samples that act as stand-ins for
parameters are called statistics – Although most analyses of data sets involve calculating statistics, remember they are only shoddy substitutes for the parameters we really want to measure Statistic: is a value usually a numerical value that describes a sample. A statistic us usually derived from measurements of the individuals in the sample |
|
Descriptive statistics
|
(10% of the course) simply
describe a data set Descriptive statistics : are statistical procedures used to summarize, organize and simplify data |
|
Inferential statistics
|
(90% of the course) allow us to go
beyond the data set and draw inferences about the meaning of the data Inferential statistics : consist of techniques that allow us to stud samples then make generalization about the populations from which they were selected |
|
Margin of error
|
is an estimate of the difference between the observed value
(sample statistic) and the true value (population parameter) ~a statistic we care about because it tells us how much confidence or trust we can have in the finding |
|
Sampling error
|
the difference between the value we have in a sample and the parameter we would like to have
- there are all kinds of reasons why our samples never completely give a perfect estimation of the population parameter Sampling error : is the discrepancy, or amount of error, that exists between a sample statistic and the corresponding population parameter |
|
Sampling error is ____ with larger samples
|
Sampling error is smaller with larger samples
– But diminishing returns – Increasing N from 100 to 1000 greatly reduces sampling error… – …adding another 1000 observations reduces it only slightly more ~ it may not be worth the additional effort and resources to make the sample larger than a certain point |
|
All statistical tests are designed to consider: (3 things)
|
– Consistency
– Magnitude of treatment effect – Magnitude of effect of random factors |
|
– Consistency
|
How frequently does the same party come out on top?
the more consistency the results have the more confidence we have in them |
|
Magnitude of treatment effect
|
How big is the difference between support for
the leading and second-place party? We want to know that the changes were large enough to be significant - the size of the difference between the two things we are comparing |
|
– Magnitude of effect of random factors
|
How much does the estimate vary from one poll to the next?
the more variability the harder it is to account for differences - the more random variability I have within a group the harder it will be to detect variability between groups |
|
Control group (condition group)
|
– A control condition provides a baseline against which
other experimental conditions can be compared – Studies often have more than one control condition |
|
Most appropriate research design depends
on two factors: |
Most appropriate research design depends
on two factors: 1) Conclusions you want to draw 2) The type of study you can conduct |
|
Two broad classes of research designs are :
|
– qualitative research designs
- quantitative research designs correlational experimental quasi-experimental |
|
Correlational Method
|
Investigate relationships that exist naturally
(no intervention from experimenter) - looking for the tendency for scores to pair up Measure two things, determine strength and direction of the relationship between them – If increase in one measurement is associated with an increase or decrease in the other measurement, the two measurements are correlated Correlational method: two variables are observed to determine whether there is a relationship between them |
|
Correlations can be either (three things)
|
Correlations can be:
Positive - an increase in one variable is associated with an increase in another variable Negative - a increase in one variable is associated with a decrease the other variable Zero |
|
Strength of correlation: measured from ________
|
Strength of correlation: measured from -1.0 to +1.0
– Correlation of +1.0 or -1.0 is a perfect correlation |