UNIT 2 COMBO of Quizlets
Index Scoring
-Decide the desirable range of index scores. -Assignment of scores for each response
Item Selection
-Face validity of items -Unidimensionality -General or specific measure -Variance
Operationalization
Range of variation (the wider the better), Degree of precision( the more detailed the better),Dimensions(as complete as possible), Defining variables and attributes(as precise as possible), Levels of measurement.
abstract
a summary of a research article. the abstract usually begins the article and states the purpose of the research, the methods used, and the major findings
systematic sampling
a type of probability sampling in which every kth unit in a list is selected for the sample. list must be random
nonprobability sampling
any technique in which samples are selected in some way not suggested by probability theory (snowballing, purposive, quota)
nonprobability sampling
any technique in which samples are selected in some way not suggested by probability theory.
Nominal
assigned to a term with censunsus
stratification
grouping of units composing of a pop from homogeneous groups (strata) before sampling, ensures small groups are included in sample. requires more info, you will know if one of the groups isn't represented
statistic
summary description of a variable in a sample, used to estimate a population parameter
What Is The Semantic Differential?
1) Technique that uses a 7-point or 5-point scale BETWEEN 2 OPPOSITE POSITIONS OR EXTREMES. Example: "Very Positive" and "Very Negative" or "Simple" and "Complex"
What Are The 5 Types Of Scaling Techniques?
1) The Bogardus Social Distance Scale 2) Thurstone Scale 3) Likert Scale 4) Semantic Differential 5) Guttman Scale
POLITICAL POLLS
LITERARY DIGEST'S (ROOSEVELT 2ND TERM) FAILURE DUE TO THE SAMPLING FRAM PROBLEM (PHONE AND AUTOMOBILE LISTS) AND GEROGE GALUP'S FAILURE DUE TO SOCIAL, RESIDENTIAL MOBILITY DURING WWII (TRUMAN'S VICTORY IN 1948)
What Two Scales Formats Data Suitable For Indexing And Scaling?
Likert and Semantic
Sampling
Selecting observations
Nominal measures
Variables those attributes have only the characteristics of exhaustiveness and mutual exclusiveness are nominal(eg., gender, religion,preference, political affiliation, college major, race, and sexual preference.)
Ordinal Measures
Variables who's attributes may be logically ranked - ordered are ordinal. (eg., social class, educational attainment,the level of importance or satisfaction and how often you date someone).
Multistage Cluster Sampling (Stratification in Multistage Cluster Sampling)
When clusters are combined into homogeneous strata, sampling error = reduced
mathematic expression
X=T+S+R where x is empirical indicator or observation
Representativeness and probability of selection
a representative sample is one that is similar to the population of interest on the characteristics the researcher is interested in. Probability of selection means that each member of the population has a known probability of being selected in the sample.
Equal probability of selection method (EPSEM)
a sample design in which each member of a pop has the same chance of being selected into the sample
Conscious and subconscious sampling bias
a sample is biased if it does not represent the population it purports to represent. A researcher may knowingly ignore certain segments of the population (conscious bias) or may fail to consider the actual characteristics and behaviors of the population, which could result in subconscious bias.
Random Selection
a sampling method in which each element has an equal chance of selection independent of any other event in the selection process.ULTIMATE PURPOSE OF SAMPLING:to select a set of elements from a population in such a way that descriptions of those elements accurately portray the total population from which the elements are selected.Probability sampling enhances the likelihood of accomplishing this aim and also provides methods for estimating the degree of probable success.
weighting
assigning different weights to cases that were selected into a sample with different probabilities of selection
operational
how a concept is measured
Cluster sampling
multistage sampling in which natural groups are sampled initially with the members of each selected group being subsampled afterward
Sample modification
occasionally the sample must be altered due to unforeseen circumstances, such as budget shortfalls
plagiarism
presenting someone else's words or thoughts as though they were your own, constituting intellectual theft
Disproportionate sampling and weighting
taking one step further, it is possible to overweight certain primary units if they are of special interest to the researcher.
sampling unit (a part of random selection)
that element or set of elements considered for selection in some stage of sampling.REASONS FOR USING RANDOM SELECTION:FIRST,this procedure serves as a check on conscious or unconscious bias on the part of the researcher.Random selection offers access to the body of probability theory, which provides the basis for estimating the characteristics of the population as well as estimating the accuracy of samples.
content validity
the degree to which a measure covers the range of meaning included within a concept
criterion- related concurrent validity
the degree to which a measure relates to some external criterion.
confidence level
the estimated probability that a population parameter lies within given confidence interval
Confidence interval
the range of values within which a population parameter us estimated to lie
Reliance on available subjects
this sampling strategy uses people (or groups, organizations, or social artifacts) that are readily accessible to the researcher
probability proportionate to size
-or PPS, this refers to a type of multistage cluster sample in which clusters are selected, not with equal probabilities but with probabilities proportionate to their size
Types of Composite Measures
Indexes Scales Typologies
Typologies
-Classification of observations in terms of their attribute on two or more variables. -Typically nominal -Typically for independent measures
Examining Empirical Relations
-Bivariate relationships- two people/items -Multivariate relationships- three or more people/items
SAMPLES
A SAMPLE CONSISTS OF A SET OF UNITS OR CASES
Sample selection
employs a systematic sample to select students.
Indexes and Scales
-Composite measures of variables -For quantitative data analysis -Ordinal Measures -To employ a rather refined ordinal measure of a particular variable
Index Validation
-Item analysis -External validation: the extent to which the construct is correlated with related constructs -Bad Index V. Bad Validators
cluster sampling
-a multistage sampling in which the natural groups are sampled initially, but the members of each selected group being sub-sampled afterwards
snowball sampling
-a non probability sampling method, often employed in field research, whereby each person interviewed may be asked to suggest additional people for interviewing
systematic sampling
-a type of probability sampling in which every Kth unit in a list is selected for inclusion in the sample--for example, every 25th student in the college directory of students
simple random sampling
-a type of probability sampling in which the units composing a population are assigned numbers
weighting
-assigning different weights to cases that were selected into a sample with different probabilities of selection
stratified sampling
(based on the second factor in sampling theory) Rather than selecting a sample from the total population,at large,the researcher ensures that appropriate numbers of elements are drawn from homogenous subsets of the population.The choice of stratification variables typically depends on what variables are available.You should be concerned primarily with those that are presumably related to variables you want to represent accurately.
EPSEM
(equal probability of selection method)a sample design in which each member of a population has the same chance of being selected into the sample.Seldom if ever perfectly represent the populations from which they are drawn.Probability sampling offers two special advantages.FIRST,probability samples although never perfectly representative,are typically more representative than other types of samples, because the biases previously discussed are avoided.probability sample is more likely than a nonprobability sample to be representative of the population from which it is drawn.SECOND,important,probability permits us to estimate the accuracy or representativeness of the sample.Probability sampling ensures that samples are representative of the population we wish to study,Probability sampling rests on the use of a random-selection procedure.
snowball sampling
(some consider to be a form of accidental sampling) a nonprobability sampling method,often employed in field research, whereby each person interviewed may be asked to suggest additional people for interviewing.This procedure is appropriate when the members of a special population are difficult to locate, such as the homeless,migrant workers,or undocumented immigrants.The researcher collects data on the few members of the target population he or she can locate,then asks those individuals to provide the information needed to locate other members of that population whom they happen to know."Snowball" refers to the process of accumulation as each located subject suggests other subjects.Because this procedure also results in samples with questionable representativeness,it's used primarily for exploratory purposes.Snowball sampling can be more than a simple technique for finding people to study,it can be a revealing part of the inquiry.
sampling frame
-that list or quasi-list of units composing a population from which a sample is selected
confidence level
-the estimated probability that a population parameter lies within a given confidence interval
stratification
-the grouping of the units composing a population into homogeneous groups (or strata) before sampling
sampling ratio
-the proportion of elements in the population that are selected to be in a sample
confidence interval
-the range of values within which a population parameter is estimated to lie
sampling interval
-the standard distance between elements selected from a population for a sample
What Is The Bogardus Social Distance Scale?
1) A Cumulative Scale. Often times criticized as too simple. 2) Bogardus Social Distance Scale is ONE type of Guttman Scale specifically deigned to study inter-group relations
What Is The Lkert Scale?
1) A measurement technique based on the use of STANDARDIZED ANSWER CATEGORIES. Example: "Strongly agree", "Agree", "Disagree", etc.
What Is The Guttman Scale?
1) A method of discovering and using the empirical intensity structure among several indicators of a given concept. Example: (Of a series of items could be) "I love Hyundai Sonata" "I am willing to buy a Hyundai Sonata" "I am willing to test drive a Hyundai Sonata" "I am willing to visit a Hyundai dealership"
What Are Inadequacies Of Single Indicators?
1) Capture all the dimensions of a concept 2) Sufficiently clear validity to warrant their use 3) Permit the desired range of variation to allow ordinal rankings
Name Two Different Types Of Variables.
1) Discrete Variables have finite attributes 2) Continuous Variables have infinite interval or ratio values, indexes and scales are mainly used for multidimensional and loosely-defined CONCEPTS and are rarely used for well-defined VARIABLES.
What Are Two Techniques That Allow Items To Be Used In An Index In Spite Of Missing Data?
1) Exclude them from data analysis 2) Use average response given by other respondents for missing data
What Does The Criteria Of Item Selection Include?
1) Face (or Logical) Validity 2) Unidimensionality (Only One Dimension of a Concept) 3) The Degree of Specificity With Which a Dimension Is To Be Measured 4) Amount of Variance (Range of Variation) Provided By The Items
What Are Indexes?
1) Index is accumulative (different parts have EQUAL WEIGHT) - just add all parts up 2) Indexes are usually Ordinal Measures, but a small percentage of them are actually Interval/Ratio Measures. 3) Most of the Composite Measures we use in daily life are Indexes 4) Indexes are based on Simple Accumulation of indicators or a variable.
What Is Item Analysis?
1) Items analysis is a type of internal validation, based on the relationship between each individual item in the Index and the Index itself. 2) Item analysis assesses whether each item makes Independent Contribution or merely Duplicates the Contributions of the Other Items. 3) If the item in question Contributes nothing to the Index's Power, IT SHOULD BE EXLUDED.
What Is External Validation?
1) Refers to the relationships between the Index and other Indicators of the same concept not included in the Index. 2) If the Index really measures the concept prejudice, for example, it should correlate with other indicators or prejudice.
What Are Scales?
1) Scales take advantage of any Logical or Empirical intensity structures that exist among a variable's indicators. 2) Scale is weighted and give DIFFERENET WEIGHTS to different components or parts. 3) ALL Scales are Ordinal Measures, unlike Indexes
What Are the 4 Principal's Steps In Constructing An Index?
1) Selecting Possible Items 2) Examining Their Empirical Relationships 3) Scoring the Index 4) Validating the Index
What Is The Thurstone Scale?
1) Technique that uses Judges to determine the intensities of different indicators (FROM THE WEAKEST INDICATOR TO THE STRONGEST INDICATOR) by giving a score (a number from 1-whatever applicable.) Example: College & University Rankings , Various Ratings
What Is Typologies?
1) Typology is a Nominal composite measure often used in Social Research. 2) Typologies may be used effectively as Independent Variables, but Interpretation is difficult when they are used as Depended Variables. Example: Types of Crime (Crime Against the Person, Organized Crime, Hate Crime, Cyber Crime) or Types of Family (Nuclear Family, Extended Family, Blended Family, Single-Parent Family)
Internal consistency reliablites
1. Cronbachs alpha -mathematically equivalent to the average of all possible split -half estimates, 2.average inter-item correlation 3.average item total correlation compute a composite measure and correlate ut with all items 4.using established measures The conflict tactic scales Gender ideology inventory 5. research - worker reliability( inter-rater reliability ) 1.clearly conceptualize contracts 2. use a precise level of measurement 3.use multiple indicators 4. use pilot tests
Nonprobability Sampling
= useful, especially in qualitative research but limitations to accurate representations of populations
SNOWBALL SAMPLING
A NON PROBABILITY SAMPLING METHOD OFTEN EMPLOYED IN FEILD RESEARCH WHEREBY EACH PERSON IN INTERVIEWED MAY BE ASKED TO SUGGEST ADDITIONAL PEOPLE FOR INTERVIEWING (E.G.,ONE OF CHILDREN: GAY BLACK MEN IN HARLEM BY WILLIAM G. HAWKERWOOD 1996)
POPULATION PARAMETER
A PARAMETER IS THE SUMMARY DESCRIPTION OF AGREN VARIABLE IN A POPULATION (E.G, THE AVERAGE OF MALES OR FEMALES IN THE UNITED STATES) PARAMETERS ARE DETERMINED WHEN ALL ELEMENTS ARE MEASURED (E.G., BY CENSUS WHICH IS AN OFFICIAL COUNT OF THE ENTIRE POPULATION AND THE RECORDING OF CERTAIN INFORMATION ABOUT EACH PERSON.
POPULATION(UNIVERSE)
A POPULATION IS THE THEORETICALLY SPECIFIED AGGREGATION OF STUDY ELEMENTS WITH GEOGRAPHICAL AND TIME BOUNDARIES .
SAMPLING
A PROCESS OF SYSTEMATICALLY SELECTING CASES FOR INCLUSION IN A RESEARCH PROJECT. A SELECTED SET OR CASES IS REFERRED TO AS A SIMPLE .
STATISIC :
A STATISTIC IS THE SUMMARY DESCRIPTION OF A GIVEN VARIABLE IN A SAMPLE.
TARGET (OR STUDY) POPULATIONS
A TARGET POPULATION IS THE AGGREGATION OF ELEMENTS FROM WHICH THE SAMPLE IS ACTUALLY SELECTED
QUOTA SAMPLING
A TYPE OF NON PORTABILITY SAMPLING IN WHICH UNITS ARE SELECTED INTO A SAMPLE ON THE BASIS OF PRE-SPECIFIED CHARACTERISTICS SO THAT THE TOTAL SAMPLE WILL HAVE THE SAME DISTRIBUTION OF CHARACTERISTICS ASSUMED TO EXIT IN THE POPULATION BEING STUDIED
PURPOSIVE OR JUDGMENTAL SAMPLING
A TYPE OF NON PROBABILITY SAMPLING IN WHICH RESEARCHERS SELECT THE UNIT TO BE OBSERVED ON THE BASIS OF THEIR OWN JUDGMENT ABOUT WHICH ONES WILL BE THE MOST USEFUL OR REPRESENTATIVE (E.G., WOMEN OF THE KLAN BY KATHLEEN BLEE, 1991)
TYPES IF RANDOM SAMPLING
A TYPE OF PROBABILITY SAMPLING IN WHICH THE UNITS COMPOSING A POPULATION ARE ASSIGNED NUMBERS , A SET OF RANDOM NUMBERS IS THEN GENERATED AND THE UNITS HAVING THOSE NUMBERS ARE INCLUDED IN THE SAMPLE
SYSTEMATIC SAMPLING
A TYPE OF PROBABILITY SAMPLING IN WHIHC EVERY kth UNIT IN A LIST IS SELECTED FOR INCLUSION IN THE SAMPLE WITH A RANDOM START.
SURVEY'S
ABOUT 18% OF ARTICLES PUBLISHED IN SOCIOLOGY JOURNALS USED THE SURVEY METHODS IN 1938-1940; THIS ROSE TO 55% 1964-65
SAMPLING FRAME
AN ACTUAL LIST OF SAMPLING UNIT FROM WHICH A SAMPLE IS SELECTED. (EX : STUDENT IDS)
SAMPLING UNIT
AN ELEMENT CONSIDERED FOR SELECTION (E.G, STREET BLOOK AS A PRIMARY SAMPLING UNIT.)
NON PROBABILITY SAMPLING
ANY TECHNIQUE IN WHICH SAMPLES ARE SELECTED IN SOME WAY NOT SUGGESTED BY PROBABILITY THEROY SUCH AS NON RANDOM SELECTION CONVENIENCE SAMPLING, PURPOSIVE OR JUDGEMENTAL SAMPLING, SNOWBALL SAMPLING,QUOTA SAMPLING, SELECTING INFORMANTS.
Scales
Advanced indexes whose observations are further transformed (scaled) due to their logical or empirical relationships
What Does Indexes And Scales Do In Social research?
Although both Indexes and Scales are intended as Ordinal Measures, Scales typically satisfy this intention better than Indexes do.
Quota Sampling (Continued)
Although quota resembles probability sampling, it has several inherent problems.First,the quota frame (the proportions that different cells represent) must be accurate,and it's often difficult to get up to date information for this purpose.(This is what caused Truman the election in 1948).Second,the selection of sample elements within a given cell may be biased even though its proportion of the population is accurately estimated.The logic of quota sampling can sometimes be applied usefully to a field research project.
Indicator
An observation that we can choose to consider as a reflection of a concept we wish to study
What Is Scale Construction?
Based on the belief that NOT all indicators of a concept are equally important, scale is a composite measure that gives different items or indicators in the scale different weights or intensities.
The Theory and Logic of Probability Sampling (Conscious and Unconscious Sampling)
Bias: Those selected are not typical or representative of the larger populations they have been chosen from (may or may not be intentional)
What Does Indexes Versus Scales Mean?
Both Indexes and Scales are composite measures of concepts or variables.
The Theory and Logic of Probability (Confidence Levels and Confidence Intervals)
Confidence level: The estimated probability that a population parameter lies within a given confidence interval Confidence interval: The range of values within which a population parameter is estimated Ex. 40% of people favour Candidate A, compute confidence interval = 35% to 45% and that in 95% of all samples the true population value will be inside the constructed interval = confidence level
Brief History of Sampling (Polling in Canada)
Cons of polls: Concerns of who and how they are conducted, and their potential impact and misuse
What Is ONE Remedy For These Inadequacies Of Single Indicators?
Create composite measures
RANDOM SELECTION
EACH ELEMENTS AS AN EQUAL CHANCE OF SELECTION INDEPENDENT OF ANY OTHER EVENT IN THE SELECTION PROCESS
The Theory and Logic of Probability (Random Selection)
Each element has an equal chance of selection independent of any other event in the selection process Ex. Flipping coin, the selection of a head or a tail is independent of previous selections of heads or tails (does not matter how many tails turn up in a row, the chance that the next flip will produce tails is always 50-50) Sampling unit: Element/set of elements considered for selection in some stage of sampling Reasons for random selection: Check on conscious or unconscious bias, estimates the characteristics of the population, estimates the accuracy of samples
Multistage Cluster Sampling (Multistage Designs and Sampling Error)
Efficient but less accurate sample Ex. 2 stage cluster = two sampling errors Typically researchers are restricted to some sample size Sampling error is reduced by two factors: Increase in sample size and increase homogeneity of the elements being sampled Therefore clusters = large number and = much alike *Typically elements making up a given natural cluster within a population are more homogeneous than are all elements making up the total population Ex. Members of a church are more alike than are all church members Therefore general guideline for cluster design is to maximize number of clusters selected while decreasing the number of elements within each cluster
If Different Items Are Indeed Indicators Of The Same Variable Then They Should Be Related To What?
Empirically to One Another
Nonprobability Sampling (Snowball Sampling)
Form of accidental sampling, each person interviewed may be asked to suggest additional people for interviewing Problem of generalization therefore mostly used for exploratory purposes Ex. Learning the pattern of recruitment and asking who introduced them to the group
SEQUENTIAL SAMPLING
GATHERING CASES UNTIL A SATURATION POINTS IS RESEARCHED (E.G., STUDYING WIDOWS OVER 75 YEARS OLD)
Brief History of Sampling (President Thomas E. Dewey)
George Gallup and quota sampling: Based on a knowledge of the characteristics of the population being sampled by selecting people to match a set of characteristics/variables that are most relevant to the study Failure: Most stopped polling in early October despite early trend towards Truman, many voters = undecided, and most importantly Gallup relied on the census data which was affected by World War 2 (movement from the country to the cities, therefore more democratic)
President Thomas E. Dewey
George Gallup had used quota sampling to predict earlier presidential winners. Because of the massive changes in the population following WW II, the quotas were no longer accurate, and he incorrectly predicted that Thomas Dewey would defeat the incumbent, Harry Truman
CENSUS
INFORMATION ON CHARACTERISTICS OF THE ENTIRE POPULATION IN A TERRITORY. (1ST US CENSUS IN 1790)
Likert Scaling
Improve levels of measurement through use of standardized response categories in surveys to determine relative intensity of different items Ex: strongly agree, agree, disagree, strongly disagree
Nonprobability Sampling (Selecting Informants)
Informant: Member of the group who can talk directly about the group, someone well versed in the social phenomenon that you wish to study and is willing to tell you what he or she knows Does not = respondent Respondent: People who provide information about themselves *Don't pick outsiders, outsiders = bias view Informants may be outsiders/marginal/atypical therefore may bias the view you get and may limit their access to different sectors you want to study
What Is Index Scoring?
Involves deciding the desirable range of scores and determining whether items will have Equal or Different Weights
Index Constructions
Item Selection Examine empirical relations Index Scoring Index Validation
Probability sampling
Key to generalizing from a sample to a larger population *Idea of random sampling
MULTISTAGE DESIGNS AND SAMPLING ERRORS
LESS EXPENSIVE THAN SRS, MORE SAMPLING ERRORS(DERIVED FROM EACH CLUSTER) THAN SRS
Populations and Sampling Frames
List/quasi-list of units making up a population from which a sample is selected, if sample wants to be representative of the population, the sampling frame then must include all (or nearly all) members of the population Ex. Sample of students is deleted from a student roster (sampling frame) *Sampling frame = population we want to study Many use telephone directories as sampling frame but need to take into account social-class bias (new subscribers, unlisted numbers, poor ppl less likely to have telephones, cellphones)
President Alf Landon
Literary Digest conducted a poll to determine the winner of the 1936 presidential election and predicted that Alf Landon would unseat President Roosevelt. They overrepresented wealthy voters who favored Landon, and thus their prediction was incorrect.
Brief History of Sampling (President Alf Landon and Literary Digest)
Literary Digest: Predict that Alf Landon = winner against President Franklin Roosevelt, but sampling frame did not represent whole population (sampling frame = telephone subscriber and automobile owners which in that time = wealthy who = + Republican) therefore inaccurate prediction
Indexes
Measures that summarize and rank specific observations, usually on the ordinal scale
Multistage Cluster Sampling (Probability Proportionate to Size (PPS) Sampling)
More sophisticated form of cluster sampling, often used in large-scale survey sampling projects Select a random/systematic sample of clusters and then a random/systematic sample of elements within each cluster selected -> every element in the whole population has the same probability of selected To calculate overall probability of a household being selected, simply multiply probabilities at the individual steps in sampling Ex. Each household has 1/10 chance of its block being selected and 1/10 chance of that specific household being selected if the block is one of the chosen then (1/10)(1/10) = 1/100 *Problem: Clusters may differ in size -> Probability proportionate to size (PPS) sampling design PPS: Guards against problem but produces a final sample in which each element has the same chance of selection Therefore each cluster is given a chance of selection proportionate to its size Ex. City block with 200 household has twice the chance of selection than one with 100 households *Within each cluster = fixed number of elements selected Ex. 5 households per bock Example: Block A = 100 households, block B = 10 houses Say A has 1/20 chance then B has 1/200 chances to be selected If Block A is selected and we're taking 5 households from each selected block, then block A has 5/100 chance of being selected into the block's sample Therefore overall chance of selection = (1/20)(5/100) = 1/400 If Block B is selected and we're taking 5 households from each selected block, then block B has 5/10 chance of being selected into the block's sample Therefore overall chance of selection = (1/200)(5/10) = 1/400 *Even though much better chance at being chosen among household, poorer chance among block therefore ends up equal
TWO TYPES OF SAMPLING METHODS
NON-PROBABILITY & PROBABILITY
Types of Sampling Designs (Stratified Sampling)
Not al alternative to random and systematic but a modification in their use Random and systematic = ensure a degree of representativeness and permit an estimate of the error present Stratified sampling: Method for obtaining a greater degree of representativeness and decreasing sampling error through organizing population homogeneous subset and select appropriate number for each Homogeneous population: Ensures that appropriate numbers of elements are drawn from homogeneous subsets of that population Ex. Stratified sample of university students = organize population by class and draw appropriate numbers (1st, 2nd... etc. year) *Can be stratified by gender, grade point, etc. 2 methods: -On the basis of the relative proportion of the population represented by a given group, select randomly/systematically several elements from that group constituting the same proportions Ex. 4.0 average 2nd year students make up 1 percent and you desire sample of 1000 students, select 10 2nd year female students with 4.0 average -Group students as describe and then put those groups together in a continuous list beginning with 1st year female students with a 4.0 average and ending with all 4th year make students with a 1.0 or below, then use systematic sample with random start from entire list
The Theory and Logic of Probability (Probability Theory)
Parameter: Summary description of a given variable in a population Ex. The mean income of all families in a city Probability theory allows to make these estimates and to judge how accurately they will represent the parameters in the population Ex. Probability theory allows researchers to infer from a sample of 2000 voters how a population of 15 million voters is likely to vote, and the margin of error *The larger the sample selected, the higher the probability of a more accurate estimate of a value of the population
Nonprobability Sampling (Reliance on Available Subjects)
Picking random people in any location does not allow control over representativeness of a sample, Problem of generalization
Multistage Cluster Sampling
Preceding section = simple procedures for sampling from a list of elements Unfortunately, sometimes social research requires selection of samples from populations that cannot be easily listed for sampling purposes Cluster sampling: Used when it's either impossible/impractical to compile an exhaustive list of the elements making up the target population, multistage sampling approach in which natural groups (clusters) are sampled initially, with the members of each selected group being subsampled afterward Ex. Select a sample of Canadian universities from a directory, get lists of students at all the selected schools, then draw samples of students from each *Repeats two basic steps: Listing and sampling
The Theory and Logic of Probability Sampling
Precise, statistical description of large population, contains samples of individuals from a population that contain essentially the same variations that exist in the population
Multistage Cluster Sampling (Disproportionate Sampling and Weighting
Probability sample = population element has a known nonzero (and equal) probability of selection Weighting: Procedure employed in connection with sampling whereby units selected with unequal probabilities are assigned weights in such a manner as to make the sample representative of the population from which it was selected *When unequal, each weight is given equal to the inverse of its probability of selection
Types of Sampling Designs (Systematic Sampling)
Probability sampling in which every kth unit in a list is selected for inclusion in the sample Ex. Every 25th student in the university directory of students *Compute k by dividing the size of the population by the desired sample size (k = sampling interval) Ex. 10,000 elements and want a sample of 1000, would sample every 10th element = k *Typically the first unit is selected at random (like in simple random sampling) *Within certain constraints, systematic sampling is a functional equivalent of simple random sampling, usually easier to do, sometimes more accurate Sampling interval: Standard distance between elements selected in the sample = population size/sample size Ex. 10 (from the preceding example) Sampling ration: Proportion of elements in the population that are selected = sample size/population size Ex. 1/10 (from the preceding example) Danger: Periodicity (arrangement of elements in the list can make systematic sampling unwise) -> bias
Types of Sampling Designs (Simple Random Sampling)
Probability sampling in which the units composing a population are assigned a number, a set of random numbers is then generated and the units having those numbers are included in the sample Seldom use simple random sampling because not feasible and not the most accurate method available *Equivalent alternative is the systematic sample (with a random start)
Nonprobability Sampling (Purposive or Judgmental Sampling)
Purposive/judgmental sampling: Select sample on the basis of your own knowledge of the population and the purpose of the study, therefore units that will be the most useful or representative Ex. Field researchers and deviant cases, use cases to better understand regular patterns
Nonprobability Sampling (Quota Sampling)
Quota sampling: Matrix/table describing characteristics of a target population with a relative proportion assigned to each cell in the matrix/table, data is collected from people having those characteristics, then weighed in proportion to the population Problem of generalization, quota frame (the proportions that different cells represents), allows for bias
CONVIENIENCE SAMPLING
RELIANCE ON AVAIBLE SUBECTS (E.G., SECOND CHANCES BY JUDITH S. WALLERSTEIN, 1989)
REPRESENATIVENESS AND PROBABILITY OF SELECTION
REPRESENTATIVENESS IS DEFINED AS THE QUALITY OF A SAMPLE OF HAVING THE SAME DISTRIBUTION OF CHARACTERISTIC AS THE POPULATION FROM WHICH IT WAS SELECTED
Brief History of Sampling (Two Types of Sampling)
Sampling based of probability theory = random sampling = far more accurate Therefore primary method for selecting large, representative samples
Brief History of Sampling
Sampling developed hand in hand with political polling Want to discover their accuracy
The Theory and Logic of Probability Sampling (Representativeness and Probability of Selection)
Reprentativeness: The aggregate characteristics of the sample closely approximate those same aggregate characteristics in the population Ex. If the population = 50% women, the sample must contain close to 50% women Question: But how close is close Representative when: All members of the population have an equal chance of being selected in the sample (EPSEM = equal probability of selection method) Samples seldom if ever perfectly represent the populations from which they're drawn 2 advantages: -Probability samples may not be perfectly representative but are more representative than other types of samples (other biases = avoided) Element: Unit about which information is collected and that process the basis of analysis, typically (but not always) in survey research elements are ppl/certain types of ppl Ex. Families, social clubs, corporations *Same as units of analysis but element = for sample selection vs. unit of analysis = data analysis Population: Theoretically specified aggregation of the elements in a study Ex. Canadian (population) = citizenship, residence, as of when (elements) Study population: Aggregation of elements from which the sample is actually selected (no guarantee that every element meeting theoretical definitions laid down will be selected therefore the redefine population under examination must be made clear)
Representativeness and Probability of Selection
Representativeness-That quality of a sample of having the same distribution of characteristics as the population from which it was selected.It is enhanced by probability sampling and provides for generalizability and the use of statistics.For our purpose,a sample is representative of the population from which it is selected if the aggregate characteristics of the sample closely approximate those same aggregate characteristics in the population.Note that,samples need not be representative in all respects,representativeness is limited to those characteristics that are relevant to the substantive interests of the study.However,you may not know in advance which characteristics are relevant.A BASIC PRINCIPLE OF PROBABILITY SAMPLING IS THAT:a sample will be representative of the population from which it is selected if all members of the population have an equal chance of being selected in the sample.
PROBABILITY PROPORTION TO SIZE (PPS)
SAMPING - A SPECIAL EFFIECIENT METHOD FOR MULTISTAGE CLUSTER SAMPLING
SAMPLING RATIO:
SAMPLE SIZE/ TARGET POPULATION SIZE
THEORETICAL SAMPLING
SELECTING CASES BASED ON INSIGHT OBTAINED IN THE FIELD
DEVIANT CASE SAMPLING
SELECTING CASES THAT DIFFER FROM THE DOMINANT PATTERN (E.G., STUDYING HIGH SCHOOL DROPOUTS)
SELECTING INFORMANTS
SOMEONE WHO IS WELL VERSE IN SOCIAL PHENOMENON THAT YOU WISH TO STUDY AND WHO IS WILLING TO TELL YOU THAT HE OR SHE KNOWS ABOUT IT (E.G., COMMUNITY LEADERS IN THE MISSISSIPPI DELTA)
Nonprobability Sampling
Social research conducted in situations that don't permit probability samples used in large-scale survey, therefore use nonprobability sampling 4 types of nonprobability sampling: Reliance on available subjects, purposive or judgmental sampling, snowball sampling, quota sampling
Selecting Informants
Someone who is well versed in the social phenomenon that you wish to study and who is willing to tell you what he or she knows about it.NOT TO BE CONFUSED WITH RESPONDENT.Are especially important to anthropologists,also important to other social researchers as well.Were they limited to specific jobs or did their information cover many aspects of the operation?This and other criteria helped determine how useful the informant might be.Select informants typical to what you are studying. Because they are willing to work with investigators,informants will almost always be somewhat "marginal" or atypical within their group.Informants' marginality may not only be biased, but it could limit their access to the different sectors of the community you wish to study.
The Theory and Logic of Probability (Sampling Distributions, and Estimates of Sampling Error)
Statistic: Summary description of a variable, used to estimate a population parameter Independent random samples will distribute around the population parameter Formula estimating how closely the sample statistic are clustered around the true value therefore allows us to estimated the sampling error Sampling error: Degree of error to be expected for a given sample design (three factors: parameter, sample size, standard error) s= √ (PxQ/n) where P and Q = population parameters for the binomial and n = number of cases in each sample Ex. 60% approval vs. 40% approval = √((0.6)(0.4)/100) *The higher the PxQ the higher the sampling error
What Are Composite Measures?
Such as scales and indexes, solve the above-mentioned problems by including several indicators of a concept or variable in one summary measure
PROBABILTY SAMPLING
THE GENERAL TERM FOR SAMPLES SELECTED IN ACCORD WITH PROBABILITY THEORY, TYPICALLY INVOLVED SOME RANDOM SELECTION
A SAMPLING BIAS:
THOSE WHO'RE SELECTED ARE NOT TYPICAL OR REPRESENTATIVE OF THE LARGER POPULATION
STRATIFIED SAMPLING
TO ORGANIZE THE POPULATION INTO HOMOGENOUS SUBJECTS AND TO SELECT THE APPROPRIATE NUMBER OF ELEMENTS FROM EACH 1. SAMPLING ERROR:REDUCED 2. METHODS OF STRATIFICATION POPULATION CHARACTERISTICS USED
interval Measures (continuous)
The first two plus the amount of distance between categories can be specified and equal.(eg, test scores,IQ scores)
The Theory and Logic of Probability Sampling
The general term for samples selected in accord with probability theory,typically involving some random-selection mechanism.Types of probability sampling include EPSEM,PPS,simple random sampling, and systematic sampling.Nonprobability sampling methods cannot guarantee that the sample we observed is representative of the whole population.Involves sophisticated statistics.If all members of a population were exactly the same,there would be no need for careful sampling procedures.FUNDAMENTAL IDEA BEHIND PROBABILITY SAMPLING IS THIS:To provide useful descriptions of the total population,a sample of individuals from a population must contain essentially the same variations that exist in the population.
Ratio Measurements
The only difference between interval and ration is that the latter has a true zero(eg., the zero exists for such variables as income and marriage rate.
Reliability
The quality of measurement method that suggests that the same data would have been collected each time in repeated observations of the same phenomenon. ~
Conceptualization
To conceptualize is to make the fuzzy and imprecise notions (mental processes) more specific and precise; an agreement about what terms mean; also called concept
research monograph
a book-length research report either published or unpublished. this is distinguished from a textbook, a book of essays, a novel, and so forth.
search engine
a computer program designed to locate where specified terms appear on websites throughout the World Wide Web
Confidence levels and confidence intervals
a confidence interval provides an estimate of the range around the population parameter, while a confidence level indicates as a percentage how confident the researcher is that the confident interval captures the true population parameter.
Dimension
a distinctive and specifiable aspect of concept (dimensions of romantic love: think:feel : behavior
Quota Sampling
a type of nonprobability sampling in which units are selected into a sample on the basis of prescribed characteristics,so that the total sample will have the same distribution of characteristics assumed to exist in the population being studied.Like probability sampling,quota sampling addresses the issue of representativeness, although the two methods approach the issue differently.Quota sampling begins with a matrix,or table,describing the characteristics of the target population.Depending on research purposes,you may need to know what proportion of the population are male as well as female.Knowing what proportions of each sex fall into various age categories,educational levels,ethnic groups,and so forth.You might also need to know what proportion is urban,eastern,male,under 25,white,working class,and the like and all possible combinations of these attributes.Once the matrix is created and assigned a relative proportion to each cell,you proceed to collect data from people having all the characteristics of a given cell.A weight is assigned to their portion of the total population.When all the sample elements are so weighted,the overall data should provide a reasonable representation of the total population.
purposive (judgmental) sampling
a type of nonprobablility sampling in which the units to be observed are selected on the basis of the researcher's judgment about which ones will be the most useful or representative.Field researchers are often particularly interested in studying deviant cases,cases that don't fit into fairly regular patterns of attitudes and behaviors inorder to improve their understanding of the more-regular pattern.Selecting deviant cases for study is another example of purposive study.In qualitative research projects, the sampling of subjects may evolve as the structure of the situation being studied becomes clearer and certain types of subjects seem more central to understanding than others do.
study population
aggregation of elements from which a sample is actually selected
contract validity
agreement between a theoretical concept (construct) and a specific measuring device or procedure
Measurement Theory
based on the idea that an empirical measure of a concept reflects three components 1. The true or idea contract or and absolutely perfect measure of it; 2. systematic errors or biases(non-random) 3. random errors, unavoidable error (change errors)
Probability proportionate to size (PPS) sampling
because certain primary sampling units have more elements than others, the elements in those large units ultimately has a smaller chance of being selected. Probability proportionate to size sampling weighs each primary sampling unit by the number of elements it has.
Multistage designs and sampling error
because multistage cluster design involves multiple probability samples, the sampling error for the final is compounded by the number of samples drawn.
conceptual order
conceptualization, nominal, operational, measurements in the field.
sampling frame
list/quasi list of units composing a population from which a sample is selected (includes all/nearly all members of pop)
sampling error
degree of error to be expected in probability sampling
An illustration: sampling university students
describes the process of sampling university studies.
Random selection
each element has an equal chance of selection independent of any other event in selection process
Simple random sampling
each element of the population is assigned a number, and a table of random numbers (now typically computers perform this task) is used to generate a random sample.
Heterogeneous Cluster
elements can be vastly different (sample heavily but not many)
systematic sampling
every kth element in the total list is chosen (systematically) for inclusion in the sample.To ensure against any possible human bias in using this method,you should select the first element at random.Is virtually identical to simple random sampling.Slightly more accurate than simple random sampling.SAMPLING INTERVAL:is the standard distance between elements selected in the sample.SAMPLING RATIO:is the proportion of elements in the population of elements in the population that are selected.
The sampling distribution of ten cases
example of drawing a sample ten times and examining the results of those ten samples. Illustrates how sampling distributions work.
Random selection
in random selection, each member or element of the population has the same chance of being sampled; the specific characteristics of the elements do not affect selection.
Implicit stratification in systematic sampling
in systematic sampling, if the list is arranged in some meaningful way, then it is already stratified and the resulting sample will be improved.
Quota sampling
in this sampling strategy the researcher knows the characteristics of the population he or she wishes to sample. The researcher then selects subjects that represent the population.
Systematic sampling
in this sampling strategy, the elements are arranged in a list and every nth element from the list is selected. N is estimated by dividing the size of the population by the size of the desired sample.
element
is that unit about which information is collected and that provides the basis of analysis.In survey research, elements are people or certain kinds of people.Other kinds of units can contribute like:families,social clubs,or corporations might be the element of study.Elements are often the same as units of analysis,though the former are used in the sample selection and the latter in data analysis.
simple random sampling
is the basic sampling method assumed in the statistical computations of social research.Once a sampling frame has been properly established,to use simple random sampling the researcher assigns a single number to each element in the list,not skipping any number in the process.Simple random sampling is not feasible and may not be the most accurate method available.
Stratification in multistage cluster sampling
it is possible to stratify the primary sampling units by important characteristics and then draw a sample from each subunit, just as in a simple stratified sample.
Levels of measurement
nominal measures, Ordinal Measures, Interval Measures, Ratio Measures
Two type of sampling methods
nonprobability and probability sampling are the two major types of sampling strategies available to sociologists. Each has its own particular advantages and disadvantages.
purposive sampling
nonprobability, units selected by basis of researchers judgement about which is more representative/useful
Quota Sampling
nonprobability, units selected on basis of pre-specified characteristics, so total sample will have same distribution of characteristics assumed to exist in the population being studied
snowball sampling
nonprobability, used in field research. those interviewed may be asked to suggest other people for interviewing, no sampling frame needed
Selecting informants
often in field research, sociologists rely on a few individuals as a source of data on the group, organization, or social phenomenon that is being examined. Because informants provide much of the information in the study, they must be trustworthy and knowledgeable.
False vaildity
on the face of it the indicators measures the construct
Probability theory, sampling distributions, and estimates of sampling error
probability theory allows researchers to estimate how close to the population their sample is on a given dimension. It is based on sampling distributions, which tell the given estimate of a population had a large number of samples been taken.
sampling ratio
proportion of elements in pop that are selected to be in a sample
Tension between reliability and validity
reliable but not valid(grades and intellectual ability valid but not reliable(# of sexual partners) reliability is necessary for validity (act and college performance)
reliance on available subjects
relying on available subjects,such as stopping people at a street corner or some other location, is sometimes called "convenience" or "haphazard" sampling. This is a common method for journalists, but it is an extremely risky sampling method for social research.This method does not permit any control over the representativeness of a sample.Justified only if the researcher wants to study the characteristics of people passing the sampling point at specified times or if less-risky sampling methods are not feasible.Even when this method is justified on grounds of feasibility, researchers must exercise great caution in generalizing from their data.They should also alert readers to the risks associated with this method.
representativeness
sample ahs some characteristics as the population from which it was selected
Sampling distributions and estimates of sampling error
sampling distributions allow the sociologists to calculate the sampling error; the amount of error made when trying to estimate a measure of the population using a sample
probability sampling
selecting in accordance with probability theory (EPSM, PPS, SRS, Systematic)
Snowball sampling
snowball sampling uses subjects as a way to identify other potential subjects to be included in the sample; it is especially useful for studying populations whose members are difficult to locate.
Stratified sampling
sociologists can increase the accuracy of their sample by dividing up the population into relevant subunits, and then selecting a random sample from each subunit, based on the portion of the population the subunit encompasses.
sampling interval
standard distance (k) between elements selected from a pop for a sample
Study population and sampling frame
the first step is defining the group of students that the researcher is interested in drawing conclusions about (the study population) and then identifying all potential students to be included in the sample (the sampling frame).
parameter
the summary description of a given variable in a population
population
the theoretically specified aggregation of the elements in a study
element
the unit of which a population is composed and which is selected in a sample
Purposive or judgmental sampling
this sampling strategy entails identifying a small subset of the population the researcher is interested in and then sampling those subjects.
probability proportionate to size (PPS)
type of multistage cluster sample in which clusters are selected not with equal probability but with probability proportionate to their size
SRS
units composing a pop are assigned numbers, generated number sets and units with those numbers are included
Homogeneous Cluster
want elements to be very similar (sample lightly but lots of clusters)
URL
web address; uniform resource locator
Stratification
what criterion should be used to stratify the sample.