Research Methods for Social Work Ch. 15 - Sampling

Purposive or judgmental sampling=

TYPE OF NON PROBABILITY SAMPLING METHOD IN WHICH THE RESEARCHER USES THEIR OWN JUDGEMENT IN SELECTING THE SAMPLE MEMBERS You select your sampling based on your own knowledge of the population, its elements, and the nature of the research 1. select units to be observed based on informed choice about which ones will be most useful (non-probability sampling) 2. YOU handpick the key people, who in your judgement, represent the range of people who will best know 3. Sometimes is use to compare OPPOSITE EXTREMES, example mental illnesses (USE IN QUALITATIVE STUDIES) ***cases that do not fit into a regular pattern of attitudes an behaviors use DEVIANT CASE SAMPLING***

Study population=

That aggregation of elements from which the sample is actually drawn 1. It has limits (national pooling might be limited to the 48 contiguous States for practical reasons)

Sampling of margin error (estimating)

The difference between the true population parameter and our estimate is the margin error (eg:TPP is 50%; our estimate if 100%, sampling error will be 50 percentage points) review tossing coin example page 363

Simple random sampling (probability sampling)

The researcher assigns a single number to each element in a sampling frame WITHOUT SKIPPING any number in the process 1. it can be selected using a random # generator ENSURES A DEGREE OF REPRESENTATIVENESS AND PERMIT AN ESTIMATE OF THE ERROR PRESENT ***you will seldom, if ever use it in practice because: sometimes simple random sampling is not feasible, and it might not be the most accurate method) is tedious if done manually, it requires a list of elements

Sample (observation unit)

The smaller group ASUBLET OF THE POPULATION) that the studies observe and make inferences about the nature of the total population itself 1. those units from a population selected and on which data is collected

Sampling error

an estimate of the error to be expected for a given sampling design

Probability theory

branch of statistics that provides the basis for estimating the parameters of a population

sampling unit =

element or set of elements considered for selection in some stage of sampling

Confidence level

estimated probability that a parameter lies within a given confidence interval

Remember

if members of a population have unequal probabilities of selection into the sample, it is necessary to assign weights to the different observations made in order to provide a representative picture of the total population. THE WEIGHT ASSIGNED TO A PARTICULAR SAMPLE MEMBER SHOULD BE INVERSE OF ITS PROBABILITY OF SELELCTION

Sampling frame

list of units composing a population from which a sample in selected

Informants=

members of a group or other people knowledgeable about it who are willing to talk about the group per se. They should be selected in a fashion as to provide a broad diverse view of the group being studied

Sampling=

process of selecting a sample of a group

SELECTING INFORMANTS IN QUALITATIVE RESERACH

qualitative research that involves an attempt to understand a social setting (gang, local neighborhood), understanding comes from the collaboration with one or more of the members of the group being studied: INFORMANTS

Confidence interval

range of values within which a population parameter is expected to lie

Availability sampling (accidental sampling, convenience sampling)=

relying on available subjects (non-probability sampling) 1. less expensive than other methods 2. because other methods are not feasible for a particular type of study or population IT CAN BE EXTREMELY RISKY as it might just represent an opinion of "where you are at" BUT it can provide useful TENTATIVE FINDINGS **valuable when it involves the use of experimental or quasi experimental designs in evaluating the effectiveness of social work practice programs (CH 11-12)-How much are your services helping particular clients, election polls

Non-probability sampling

sample selected by other than probability theory 1. Social research is often conducted using non-probability sampling 4 TYPES of NON PROBABILITY SAMPLING PROCEDURES 1. Reliance on available subjects 2. purposive or judgmental sampling 3. quota sampling 4. snowball sampling

Quota sampling= (George Gallup 1936 & 1948 "disasters on president's predictions")

selecting a sample that has the same distribution of characteristics as the population (non-probability sampling) 1. begins with a matrix that describes the target population characteristics (eg: male, female, age) 2. proportions are assigned to each cell(how many are urban, rural, etc) 3. data is collected from people who has all the characteristics of the given cell WEIGHING:-ASSIGNED PROPORTION OF THE TOTAL POPULATION PROBLEMS: 1. Quota frame( the proportion that different cells represent) must be accurate 2. Biases might exists in the selection of the sample elements within a cell

Statistical power analysis

statistical procedure to guide decisions about sample size. 1. IT deals with determining how large a sample needs to be in order for researchers to have an adequate probability of obtaining statistical significant findings eg: sample power p.365 ***decisions about sample size in social work research rarely involve estimating sampling error (often due to practical limitations)

Parameter

summary description of a given variable in a population

Statistic

summary description of a given variable in a sample

parameter

the summary description of a given variable in the population example: the mean income of the families in a city is a parameter; the age distribution of the city population is a parameter

Element

the unit about which information is collected and that provides the basis for analysis 1. in a survey research the elements are people or certain types of people 2. it can also be: families, social clubs, corporations 3. Sometimes they are referred as UNITS OF ANALYSIS

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

Multistage cluster sampling(probability sampling) p372 for examples see p.374 for an illustrative example on cultural diversity grant

used when sampling frame unavailable; when is impossible or impractical to compile an exhaustive list of elements that compose the target population involves repetition of two basic steps -- listing and sampling 1. the list of primary sampling units is compile, and perhaps stratifies; and then a sample of those units is selected. The selected primary sampling units are listed and perhaps stratified and then sampled...

Nonresponse bias=

when a substantial proportion of the people in the randomly selected sample chose not to participate in the study

Stratified sampling(probability sampling)

**NOT an alternative to random or systematic but a possible modification Method for obtaining a greater degree of REPRESENTATIVENESS-FOR DECREASING THE PROBABLE SAMPLING ERROR 1. units in a population are grouped before sampling and the appropriate number of elements are drawn from homogeneous subsets of that population 2. ULTIMATE FUNCTION is to organize the population into homogeneous subsets (with heterogeneity between subsets) and to select the appropriate # of elements from each

Systematic sampling (probability sampling)

-------FIND DEFINITION ENSURES A DEGREE OF REPRESENTATIVENESS AND PERMIT AN ESTIMATE OF THE ERROR PRESENT sampling interval-standard distance between the elements selected in the sample sampling ratio-the proportion of the elements in the population that are selected DANGER INVOLVED IN SYSTEMATIC SAMPLING: the arrangements of elements in the list can make systematic sampling unwise (PERIODICITY-IF A LIST OF ELEMENTS IS ARRANGED IN A CYCLICAL PATTERN THAT COINCIDES WITH THE SAMPLING INTERVAL, a GROSSLY BIASED SAMPLE MAY BE DRAWN 1. is usually superior to simple random samples (convenience) 2. Be aware of biases in choosing the order of elements to avoid bias

Randomly selected samples can be biased if

-mistakenly population is equate to any list of elements from which the sample is selected -is assumed that the intended sample has been randomly selected BOTH RELATED TO SAMPLING FRAMES

It is important to take steps to avoid...

....sampling bias based on race, class, gender and other significant characteristics.

Representativeness and probability of selection(p358)=

A sample is representaive of its population if the sample's aggregate characteristics closely approximate those same aggregate characteristics in the population BASIC PRINCIPLES OF PROBABILITY SAMPLING 1. a sample will be representative of its population if all the members of that population have an equal chance of being selected in the sample. (NON ZERO PROBABILITY OF BEING SELECTED) a.. the samples are often labeled: EPSEM (equal probability of selection method) 2. REMEMBER EPSEM samples, seldom, if ever represent the population for which they are drawn ...BUT IT HAS 2 SPECIAL ADVANTAGES: a. Probability samples, even if never perfectly representative, are typically more representative than others, because BIASES are avoided b. Probability theory permits us to estimate the sample accuracy or representativiness (success vs. failure)

Snowball sampling=

Appropriate when members of a special population are difficult to find (migrant workers, undocumented immigrants) 1. Sometimes considered a form of "accidental sampling"sample 2. selected by asking each person interviewed to suggest additional people for interviewing (non-probability sampling) 3. USED PRIMARILY f=FOR EXPLORATORY PURPOSES & QUALITATIVE RESEARCH

Probability sampling (the logic)+

If the sample of individuals from a population is to provide useful descriptions of the total population, then it must contain essentially the same variations that exixts in the population 1. A sample selected according to probability theory

disproportionate Stratified samples

In a disproportionate stratified sample the population of sampling units are divided into sub-groups, or strata, and a sample selected separately per stratum. Crucially, the sampling fraction is not the same within all strata: some strata are over-sampled relative to others.

Proportionate Stratified samples

In a proportionate stratified sample, the population of sampling units are divided into sub-groups, or strata, and the sample is selected separately in each stratum. For the sampling to be proportionate, the sampling fraction (or interval) must be identical in each stratum.

Representativeness=

In quantitative descriptive and explanatory studies, a key issue is whether the sampling procedures used are likely to yield a study sample that is likely representative of the larger population which the study seeks to generalize 2 CRITERIA: ***size of the sample(adequate size) & if it was selected in an unbiased way**** 1. A sample will be representative if all members of the population have an equal chance of being included in the sample; sample has same characteristics as the population

Population

Is the theoretical specified aggregation of study elements; that group about whom the researcher would like to make generalization

Random selection=

Key to the element of sampling 1.each element has an equal chance of selection that is independent of any other event in the selection process example: flipping a coin reason for using random selection method 1. this procedure serves a check on conscious or unconscious bias on the part of the researcher; erasing that danger 2. Offer access to the body of probability theory, providing the basis for estimating the characteristics of the population as well as estimates of the accuracy of samples

SAMPLING FRAMES AND POPULATIONS

Sampling frame is the list of elements from which a sample is selected 1. Properly drawn samples provide information appropriate for describing the population of elements composing the sampling frame-nothing more 2, the sampling frame must be consonant with the population we wish to study example: studies of organizations (because you have a membership list which is the perfect sampling frame), telephone directories, tax maps