Chapter 4 Sampling, Measurement, and Hypothesis Testing Flashcards
Sample
A portion or subset of a population
Population
All of the members of an identifiable group
Probability Sampling - Representative Sample & Biased Sample
- probability sampling, each member of the population has a definable probability of being selected for the sample.
- an entire population is seldom tested in a study, the researcher hopes to draw conclusions about this broader group, not just about the sample. Thus, it is important for the sample to reflect the attributes of the target population as a whole. When this happens, the sample is representative; if it doesn’t happen, the sample is biased.
- representative = A sample with characteristics that match those attributes as they exist in the population.
- biased sample = A sample that is not representative of the population.
Simple Random Sample
- A probability sample in which each member of the population has an equal chance of being selected as a member of the sample
- Simple random sampling is often an effective, practical way to create a representative sample.
Stratified sample
A probability sample that is random, with the restriction that important subgroups are proportionately represented within it.
- example, with a goal of a sample of 100, 60 women would be randomly sampled from the list of female students, and 40 men would be randomly selected from the list of male students. Note that some judgment is required here; the researcher must decide how many layers (or strata) to use. In the case of abortion, women and men were sampled in proportion to their overall numbers.
- ex provinces are strat groups and within each strata is the city
Cluster sample
- the researcher randomly selects a cluster of people all having some feature in common.
- A probability sample that randomly selects clusters of people having some feature in common (e.g., students taking history courses) and tests all people within the selected cluster (e.g., all students in three of the nine history courses available).
- example Suppose you wanted to find out how students liked living in the high‐rise dorms on your campus, which you’ve defined operationally as any dorm with eight floors or more. Further suppose that 15 of these buildings exist on your campus, housing a total of 9,000 students. Using cluster sampling, you could first select six of the buildings (each building = one cluster), and then, for each building, randomly select three floors and sample all of the residents (about 40 per floor, let’s say) of the selected floors in the selected dorms.
Non-Probability Sampling (hint 4)
- Convenience sample = A non‐probability sample in which the researcher requests volunteers from a group of people who meet the general requirements of the study (e.g., teenagers); used in most psychological research, except when specific estimates of population values must be made.
- Two other forms of convenience sampling are quota sampling and snowball sampling.
- Quota sample = A non‐probability sample in which the proportions of some subgroups in the sample are the same as those subgroup proportions in the population. same goal as stratified sampling—representing subgroups proportionally—but does so in a nonrandom fashion
- Snowball sample = A non‐probability sample in which a member of a particular group, already surveyed, helps recruit additional group members through a network of friends; often occurs for surveys of a relatively small group or a group that generally wishes to remain hidden.
-Purposive sample = A non‐probability sample in which the researcher targets a particular group of individuals (e.g., Milgram using working adults and avoiding college students).
Evaluating Measures (hint 2)
Determining if a measure is any good requires a discussion of two key factors: reliability and validity.
Reliability and Measurement Error
- Reliability = The extent to which measures of the same phenomenon are consistent and repeatable; measures high in reliability contain a minimum of measurement error.
- Measurement error = Produced by a factor that introduces inaccuracies into the measurement of some variable. (variability in your measurement)
- you can’t have an unreliable measurement and a valid measurement
Valid and all the types
Validity = In general, the extent to which a measure of X truly measures X and not Y (e.g., a valid measure of intelligence measures intelligence and not something else).
- Content validity = Occurs when a measure appears to be a reasonable or logical measure of a trait (e.g., as a measure of intelligence, problem-solving has more content validity than hat size). Content validity also concerns whether the measure includes items that assess each of the attributes.
- Face validity = Occurs when a measure appears, to those taking a test, a reasonable measure of some trait; not considered by researchers to be an important indicator of validity.. which is not actually a “valid” form of validity at all
- Criterion validity = Form of validity in which a psychological measure is able to predict some future behaviour or is meaningfully related to some other measure.
- Criterion validity is further subdivided into two additional forms of validity: predictive validity and concurrent validity
- Predictive validity = A form of criterion validity in which a measure can accurately forecast some future behaviour.
- Concurrent validity = A form of criterion validity in which a measure is meaningfully related to some other measure of behaviour.
- For example, for a test to be a useful intelligence test, it should (a) do a reasonably good job of predicting how well a child will do in school and (b) produce results similar to those produced by other known measures of intelligence behaviour. - In the examples above, the criterion variables are (a) future grades in school (predictive) and (b) scores on an already established test for intelligence (concurrent).
- Construct validity = In measurement, it occurs when the measure being used accurately assesses some hypothetical construct; also refers to whether the construct itself is valid; in research, refers to whether the operational definitions used for independent and dependent variables are valid. - relates to whether a particular measurement truly measures the construct as a whole.
- but construct validity research includes two additional procedures: convergent and discriminant validity.
- Convergent validity = Occurs when scores on a test designed to measure some construct (e.g., self‐esteem) are correlated with scores on other tests theoretically related to the construct.
- Discriminant validity = Occurs when scores on a test designed to measure some construct (e.g., self‐esteem) are uncorrelated with scores on other tests theoretically unrelated to the construct.
- high convergent and discriminant validity then its construct validity is strengthen
Measurement Scales (hint 4)
Ways of assigning numbers to events; see Nominal, Ordinal, Interval, and Ratio scales.
Nominal Scale
Nominal scale = Measurement scale in which the numbers have no quantitative value, but rather identify categories into which events can be placed.
-categorical like fruit
Ordinal Scale
Ordinal scale = Measurement scale in which assigned numbers stand for relative standing or ranking
-categorical like ranking of players
Interval Scale
Interval scale = Measurement scale in which numbers refer to quantities and intervals are assumed to be of equal size; a score of zero is just one of many points on the scale and does not denote the absence of the phenomenon being measured.
Ex. Temperature in Celsius
the difference between 45C and 35C is the same as the difference between 20C and 10C, we have equal intervals
-also 0C does not mean the absence of heat, so there is no true zero point
their difference between 140 cm and 130 cm is the same as the difference between 180cm and 170cm
-also, 0cm means a lack of height so there is a true zero point
Ratio Scale
Ratio scale = Measurement scale in which numbers refer to quantities and intervals are assumed to be of equal size; a score of zero denotes the absence of the phenomenon being measured.
Ex. Height
