week 8 quantitative methods : measurement and sampling Flashcards
what do measurements do?
link data to concepts
why is studying measurement important?
helps understand the type of data collected, understanding appropriate methods and better research
what is conceptualization?
- understanding abstract concepts
a definition in abstract, theoretical terms
“the process of thinking through the various meanings of the concept”
EXAMPLE: “CRIMINALITY”
A study is interested in people’s “criminality”
what is conceptual definition and the operational definition
Conceptual definition
* “Non-sanctioned acts of violence against other members of society or their property”
Operational definition
* Counting a person’s number of criminal arrests from official records; OR
* Calculating the amount of time a person has spent in prison; OR
* Asking people whether they have committed crimes
what is operationalization?
development of research procedures that will give empirical observations
why are operational definitions important in research?
- force us to think carefully
- allow replication
- measureable
example * What is a conceptual definition of ethnicity?
* What is an operationalized definition of ethnicity?
conceptual : * Ethnic identity which refers to self-identification within a particular group
* Ethnic origin which refers to classification based on the ethnic group to which the individual’s ancestors belong
operational : Ethnic identity:
* The individual’s self-identification within specific ethnic groups Common survey question:
* “Please write down the term that best describes the ethnic character of your everyday home environment”
Ethnic origin:
* Used in many national surveys (National Population Health Survey, Canadian Community Health Survey).
* Country of birth
* Language
categorical vs continuous variables
continouous : you take an average and it has meaning : eg: age or blood pressure - changing
categorical variable : you measure and it has no meaning - average smoker
what are discrete variables?
- variable that can only take on a certain number of values
are categorical variables considered to be discrete variables?
yes! - only certain number of values
are the following categorical or continuous
1. age
2. achievement test : pass/fail
3. score on achievement test : 1-100
- continuous
- categorical
- continuous
what is the nominal level of measurement ?
- allows researchers to classify characteristics of study pop into categories
- least precise
- no math
whats an example of nominal measures
blood type - only one category per person
what is your occupation? - choose one
what is the ordinal level of measurement
- categories and categories can be ordered in a meaningful way
- rank accoringly to charactertistics and object posses
- mutally exlusive
what is an example of ordinal measurement?
SES
- low
-medium
- high
how would u describe your health?
- poor
-good
-excellent
highest level of schooling
- elementary
- high school
-university
what is the interval level of measuremnt?
- ranked in order and actual value between values has some meaning
- numbers have meaing but no true zero point eg:: tempature
what are ratio levels of measurement?
all the other charactertistics and also have an absolute zero point which represents absence
- fixed measuring units
example of ratio levels
- body weight
- zero kg = no weight
- person who is 20kg is twice as heavy as 10kg
what is each of the level of measurement for : John is 10 years old and Sam is 20 years old
1. nominal
2. ordinal
3. interval
4. ratio
- john is young and sam is old
- john is younger than sam
3.john is 10 years younger than sam - sam is twice as old as john
what is reliability
the ability of a measuring instrument to produce consistent results under similar conditions
what is test-retest reliability?
reliability across time
what is inter-rater reliability?
independent evaluations conducted by different individuals
what is parallel forms of reliability?
reliability across indicators (i.e., two versions of the same scale should yield nearly identical results)
what is internal consistency?
Whether different items on the same test correlate
Degree to which scales are correlated, made on the assumption that scales should measure a single construct
how to improve reliability?
- Conceptualization
- Increase level of measurement i.e. use a ratio instead of an ordinal variable
- Multiple indicators
what is validity?
The degree of confidence we can place on the inferences we make about people based on their scores from that scale”
* Is the scale actually measuring what we think it is
what is the Construct validity
construct: Theoretical concept, theme or idea based on empirical observations. It’s a variable that is not directly measurable. It may consist of multiple dimensions.
what are examples of construct validity?
Example: social anxiety
Psychological dimension: intense fear and anxiety Physiologic dimension: physical stress response Behavioural dimension: avoidance of social settings
what is the construct of validity
- how well a test measures the concept it was designed to evaluate
- If the measure is valid, do the various indicators operate in a consistent manner?
what is Convergent validity:
Used for multiple indicators based on the idea that indicators of one construct will act alike or converge.
discriminant validity
Used for multiple indicators based on the idea that indicators of different constructs diverge.
Can a test truly be valid if it is not reliable?
No. A test CANNOT be truly valid if it’s unreliable.
* If the measure is not reliable, that means it gives different results every time it is
assessed.
* If it keeps giving different results, it cannot be measuring what you think it is.
what is a sample
group of indvdiuals chosen to represent a larger population
qualitative sampling
- less focus on representativeness
- focus on relevance of sample to research topic
- cases that will enhance what researchers learn
- non- probability samples
quantitative samples
- representativeness
-produce accurate generalizations about larger group - probability samples
what is non-probability sampling?
participants selected based on their relevance to the research topic rather than their representativeness
issue : not generalizable
why use non-probability samples?
- less $$
- Participant engagement can be challenging with stigmatized behaviours
- Difficult to obtain large sample of rare groups
- Difficult to capture “hard-to-reach” populations
haphazard sampling
convience sampling
quota sampling
Interviewers told they need to go out and get a given “quota” of subjects
purposive sampling
- Participants selected for reasons linked to the research study
- Typically difficult-to-reach populations
snowball sampling
- Identify a few key individuals
- Ask them to distribute
questionnaire to/recruit others - Goal is to capture an already- existing network
judgement sampling
Purposeful selection of a “representative” sample
Target population: (N)
Target population: (N)
The concretely specified large group of many cases from which a researcher draws a sample and to which results from a sample are generalized
The population of interest about which inferences are desired
sampling frame
list of all avaible sampling units in the target population- phone list, school, drivers licence
sampling ratio
The ratio of the size of the sample to the size of the target population
what is probability sampling
A method of sampling that allows inferences to be made about the population based on observations from a sample
two criteria of probability sampling
- random
- must have a non-zero chance of being selected
parameter def
True characteristic of the population
statistic def
Information from the sample to estimate a population parameter
what you actually observe in the data
benefit of probability sampling
avoids selection bias
allows generalizability
sampling error def
the degree that sample deviates from a population
what do we mean by random?
Random refers to a selection process that gives each element/unit in a population an equal probability of being selected
types of probability sampling
Simple random sampling
* Systematic sampling
* Cluster sampling
SRS simple random sampling pros and cons
Advantages
* Simple to conduct
* Disadvantages
* Requires list prior to sampling
* Can be expensive
what is systematic sampling?
- a gap or interval between each selection eg: every third
systematic sampling interval
Tells the researcher how many elements to skip in the sampling frame before you pick one of your sample.
* To calculate you need: 1) the sample size
2) population size
* Example: If a systematic sample of 300 patients were to be carried out in a family practice with a total practice population of 3,000, the sampling interval will be:
* N/n = 3,000/300 = 10
what is stratified sampling
divided pop into groups that differ in important ways
select random sample within each group
each group is mutally exclusive
what is cluster sampling?
The entire population is divided into clusters or groups and a random sample of these clusters is selected
* Typically used when the researcher cannot get a complete list of the members of the population they would like to study, but they can get a list of groups
stratification vs clustering
Stratification
* Divide population into groups (“strata”) that are different from each other
* Randomly sample from each group
Clustering
* Divide population into comparable groups
* Randomly sample some of the groups
does a large sample guarantee a representative sample
no
