discussion board questions/ experimental design activity

missruf

VSUchap6ResearchCh_6.ppt

Home >Law homework help >Criminal homework help >discussion board questions/ experimental design activity

Sampling

Chapter 6

Introduction

Sampling is the process of selecting observations

Often not possible to collect information from all units you wish to study

Often not necessary to collect data from everyone out there

Allows researcher to make a small subset of observations and then generalize to the rest of the population

The Logic of Probability Sampling

Samples: a group of subjects selected from a population

Probability sampling: a method of selection in which each member of a population has a known chance of being selected

Enables us to generalize findings from observed cases to a larger unobserved population

Eg. interviewing a sample of community residents, we may generalize findings to all community residents

Because we are not completely homogeneous/identical, our sample must be representative of the variations that exist among us

Conscious and Unconscious Sampling Bias see pg. 142, fig. 6.2

Be conscious of bias –

Bias/partial means that those selected are not typical or representative of the larger population from which it was selected

Ex. Selecting 70% women in a micro-population of 50% women and 50% men; not selecting any African American where they represent 12% of the population

Personal leanings or biases may affect sample selected thru convenient method

Eg. a researcher may avoid interviewing prosperous lawyers because of the fear of being ridiculed

Representativeness and Probability of Selection

A sample is representative of the population from which it is selected if the aggregate/total characteristics of the sample closely approximate the same aggregate characteristics in the population

Eg if the population contains 50% women, a representative sample will also contain almost 50% women

Samples that are representative of the population are often labeled equal probability of selection method (EPSEM) samples

because all members of the population have an equal chance of being included in the sample

Sampling Terminology 1

Sample Element: who or what are we studying (eg. student, court cases), unit of analysis

Population: whole group (college students – full time & part time s)

Population Parameter: summary description of a given variable in a population eg. the average income of all families in a city

Sample Statistic: summary description of a given variable in a sample;

we use sample statistics to make estimates or inferences of population parameters

Eg. average income computed from a sample are statistics & those statistics are used to estimate income parameter in a population

Sampling Terminology 2

Sampling frame: actual list of elements/units to be selected (our school’s enrollment list). The elements are the individual students

Standard error: a measure of sampling error; we can estimate the degree to be expected

The standard error is a function of the sample size i.e as the sample size increases, the standard error decreases

The difference between what is observed (sample) and what is expected or the actual (population)

Probability Sampling Designs

Simple Random Sampling:

SRS assumes unbiased sampling

the researcher assigns a single number to each element in the list, not skipping any number in the process

Then a table of random numbers or a computer program for generating them, is then used to select elements for the sample

Probability Sampling Designs

Systematic Sampling (SS):

In SS, the researcher chooses all elements in the list for inclusion in the sample

If a list contains 10,000 elements, we want a sample of 1,000, we select every tenth element for our sample

To avoid bias, we select the first element at random,

we may begin by selecting a random number b/w 1 and 10, then plus every 10th element following it

Probability Sampling Designs

Stratified sampling/grouping: modification to random and systematic sampling;

It is a method for obtaining a greater degree of representativeness

It decreases the probable sampling error

Rather than selecting from the total population, we select elements from homogenous subsets of that population

Eg. To get a SS of Uni. students we 1st organize our population by college class eg freshmen, sophomore, junior and seniors

In a SS by college class, the sampling error on that variable is almost reduced to zero

Probability Sampling Designs

Disproportionate stratified sampling:

Purposely produce samples not representative of a population on some variables

If only a small number of people in a population exhibit some characteristic of interest or the population vary widely on some variable

Then a large sample must be drawn to produce adequate numbers of elements that exhibit that uncommon condition

DSS is a way of obtaining sufficient number of rare cases by selecting a number disproportionate to their representation in the population

Eg the NCVS see next page

National Crime Victimization Survey

Seeks to represent the nationwide population of crime victims

Because crime victimization for certain offenses such as robbery or aggravated assault are rare on a national scale

Therefore, persons who live in large urban areas where serious crime is more common are disproportionately sampled

Therefore larger samples must be drawn to produce adequate numbers of elements that exhibit the uncommon condition

Multistage Cluster Sampling

Cluster sampling is used when it is impossible to compile an exhaustive list of elements that compose the target population

Ex. The pop of a city or state and all police officers in the U.S.

CS involves initial sampling of groups of elements-clusters-followed by the selection of elements within each of the selected clusters/groups

Ex. Sampling city’s population –

You might sample city blocks, list the household on each selected block, then sample the households, list the persons who reside in each household and finally sample persons within each selected household

Nonprobability Sampling

Nonprobability Sampling: the likelihood any element will be included in the sample is unknown

eg studying auto thieves; there is no list for all auto thieves

Nonprobability Sampling

Purposive sampling: selecting a sample on the basis of your knowledge of the population, its elements & the nature of the study

Based on the judgment and purpose of the study

Eg. we may wish to study a small subset of a larger population in which many members of the subset are easily identified but not all of them

Eg. Studying members of a community crime prevention groups,

many members are easily visible but it is not possible to sample all members,

we collect data from the visible members sufficient for our purposes

Nonprobability Sampling

Quota sampling: Based on the proportion of the population characteristics

Units are selected so that total sample has the same distribution of characteristics as are assumed to exist in the population being studied

It begins with a matrix/table describing the characteristics of the target population

You need to know the proportion of population that comprises; male/female, age categories, education levels

We can then collect data from people who have characteristics of a given cell

It Addresses the issue of representativeness

Nonprobability Sampling

Snowball sampling:

Commonly used in field observation studies

It begins by identifying a single subject or small number of subjects and then ask them to identify others like them who might be willing to participate in the study.

The researcher may make an initial contact by consulting criminal justice agency records to identify say someone convicted of auto theft and placed on probation

The person is interviewed and asked to suggest other auto thieves etc

Example: Snowball Sampling

To study cannabis users, Hammersley and Leon (2006) gathered a snowball sample of 176 University students who had used marijuana at least once. Extensive interviews were then conducted with the University students in the sample. Their results showed that there were two types of users—those who used cannabis on a regular basis and those who used cannabis on occasion. The results also showed that users experienced both positive and negative effects from using marijuana and the patterns of use were more similar to patterns of alcohol and tobacco use than to patterns of controlled substance use.

Hammersley, R. & Leon, V. (2006). Patterns of cannabis use and positive and negative experiences of use amongst university students. Addiction Research and Theory, 14(2), 189-205.

Nonprobability Sampling

Reliance on Available Subjects or Convenience Sampling or “person on the Street Sampling

Stopping people on a street corner or other locations

University researchers frequently conduct surveys among students enrolled in large lecture classes

It is easier, less time & inexpensive

It seldom produces data of any general value