Home > Preview
The flashcards below were created by user
on FreezingBlue Flashcards.
The fundamental research methods procedure. 5 parts
- Develop your hypothesis
- Data collection
- Selection of analysis methods
- Draw Conclusions
- Report Findings
Data is single or plural
False. Data is always plural.
Always do a _______ and inferential analysis when selecting analysis methods
Define Data for statistical analysis
Facts and statistics collected together for reference or analysis
How do you describe an individual 'piece' of data?
Each piece of data is an element under study. Data is made up of a number of observations.
What is 'population'? in research?
the total 'set' of things under investigation.
What is a sample? What does it represent?
A random selection from a population, meant to represent the total population.
What is sampling error?
A sample that is not representative of the total population.
Give an example of sampling error.
Population is entire class height average. Sample only the tallest members of the class. Results skewed.
What is non-sampling error? 5 parts
- human error
- bias with participants
- using wrong
- instrument to acquire info
- no response set in survey
What is the difference between a statistic and a census?
- Statistic a value you get from the sample.
- Census is acquiring information from the entire population.
What is a descriptive statistic, give example?
A measure that will describe the data, i.e. mean, standard deviation etc.
What is an inferential statistic?
A conclusion drawn from the sample about your population.
Give an example of a constant variable.
- Quantitiy or factor that doesn’t change between
- samples, objects, data sets. I.E. speed of light
Give an example of a variable in quantitative and qualitative.
Hair colour, temperature
Define quantitative (2 pt).
What kinds of quantitative? (2 pt.)
- Something that can be measured in numbers
- those numbers have DEFINITE meaning
Discreet or continuous.
Whole number. Fixed set of possible outcomes
Variables that can take any value within a range.
Give an example of a discreet and a continuous variable
Rainfall in March. The number of immigrants in Canada in 1987.
What are the four levels of measurement?
- Nominal Scale
- Ordinal data
Define nominal scale
It has numbers, but they have no meaning. They cannot be used in mathematics. I.E. Giving a province a number. Hair colour. Can't be ranked.
Define ordinal data
- data can be ranked, but you CANNOT do math with it.
- You can reduce ordinal data to nominal.
Define interval data
- Data has definite meaning. You can do math with it
- Interval has a relative zero and therefor can be in the negatives.
Define ratio data
Data has a definite meaning. You can do math with it.
Data has a DEFINITE zero can cannot have negatives.
Define Percision and Accuracy.
- How often you can get a value
- after repeating the measurement a number of times.
- How close you are to the actual
- value. Measure of the extent of system wide bias in your measurement.
What are the four decisions you must make for research design?
What information will you generate? NEED WELL DEFINED HYPOTHESIS
Methods of Data Collection
Coverage of Data
How to analyze the data
What is the central goal of sampling?
To derive a truly representative set of values from a population
What steps are there to define a sampling survey? (7)
- State objectives of the survey
- (hypothesis must explicitly state).
- Define target population.
- Define data to be collected
- Define the required precision
- and accuracy. This will help to define the sample size.
- Define the measurement
- instrument. (survey, observation? How does that work? What’s the best way of
- doing it?)
- Define the sample frame, sample
- size, sample method, then select the
- actual sample.
What quality should a sample size have?
sample size the minimum sample you should obtain for your study
What are the advantages of sampling over a census? (4 pts).
- Efficient and cost effective
- Can get more details from sample
- high detailed information
- high detailed accuracy
What are disadvantages to sampling? What would these examples be called?
- Selecting improper sampling
- design (are the variables
- appropriate or not? Are they the best way of achieving the end goal?)
- Method of data collection may
- be inappropriate
- Inconsistences in operational,
- logistic or personelle
Define sampling error
The amount of error you contain in your sample.
Define imprecision in sampling
Imprecision means that your example is not representative of your population
The larger the _____, the closer to the _____ you become and there will be lower _____.
N, census, impercision
In sampling you want to find the sample size
where you balance _____ and ______
Finish these sentences:
Who do you want to generalize to?
What population can you get access to?
How can you get access to them?
Who is in your study?
- Theoretical population
- Study Population
- Sampling Frame
- The Sample
Define target population:
Target population-complete set of individuals from which information is to be collected
Define target area.
Target area – entire region or set of location from which information is to be collected
Target Area: There are _____ wrong ways of defining it. There are _____ right ways of defining it.
How do you create a sampling frame?
- Define the target population into sampling units.
- Create a finite list of sampling units that make up the target population
What is the implication if a sample frame is incorrectly defined?
It will be unrepresentative of the target population.
What ways can a sample frame be inaccurate? (3)
- contains too many individuals
- contains too few individuals
- contains the wrong individuals
What are two inherent problems with sampling frame?
- Can't sample from mobile populations
- Statistical population not known
When can you not use probability methods?
When it is on humans.
Define non-probability methods. How is it made to be representative?
- Cannot specify a probability for selection. People can refuse. Trees cannot.
- Researcher uses expertise to ensure representative sample.
What are three types of non-probability sampling?
What are three problems with non-probability sampling?
- May fail to secure representative sample
- Method may make framework more difficult
- Valid inferences can't be made to larger pop'n. You can't assign probability to the sampling frame.
Define probability sampling
Every unit has an equal chance of being selected.
How do you systematically sample?
Select an r randomly, and then select a k interval.
What is a stratified sample and when is it used?
You have two or more sub-populations with important yet different features (i.e. eye correction vs. none). Split up both groups and then reassign probabilities to each group.
Define proportional stratified sample
Each group has equal probabilities (equal numbers).
Define disproportional sampling
Each group has different probabilities (and therefor different proportions)
Define cluster/area sampling
an appropriate number of categories is selected for detailed analysis through random sampling
i.e. 9 neighbourhoods in town, randomly sample 2.
When would you use spatial sampling?
When you have a map and continuously distributed variable across this map
How do you randomly select for spatial sampling?
Must randomly select X and Y.
Give an example of systematically spatial sampling
Pick an X and Y point, then a distance between each sample, and randomly select a repetition. It will result in an orderly sequence of sample plots.
Give an example of stratefied point sample
Randomly break up the map into a quadrant and then select points within each quadrant randomly