Upcoming SlideShare
×

Like this presentation? Why not share!

# Bpp week 7 sampling methods

## on Nov 25, 2011

• 1,648 views

Overview of key concepts and methods of samlping

Overview of key concepts and methods of samlping

### Views

Total Views
1,648
Views on SlideShare
1,626
Embed Views
22

Likes
0
52
0

### 2 Embeds22

 https://bpp.blackboard.com 21 https://www.edu-apps.org 1

### Report content

• Comment goes here.
Are you sure you want to

## Bpp week 7 sampling methodsPresentation Transcript

• International Pre-Masters Diploma in Business Studies.International Pre-Masters Diploma in Legal Studies.Project & Report Session 7: Introduction to Sampling methods Liam Greenslade Professional Education: developing your career
• Learning Objectives Explain the role of sampling in the research process Distinguish between probability and nonprobability sampling Understand the factors to consider when determining sample size Understand the steps in developing a sampling plan
• Formulation of the Research Project A. Selecting the Appropriate Method B. Selecting the Participants C. Selecting Measurement Methods & Techniques D. Selecting Instrumentation
• Sampling Sampling is the process of selecting a small number of elements from a larger defined target group The information gathered from the small group will allow judgments to be made about the larger groups We aim to obtain samples which are representative So that our data is generalisable
• What is a sample? A sample is a segment of the population we wish to study A great deal of care is needed in selecting the sample so that we can say that the data we obtain from it is meaningful For these reasons social scientists have a number of sampling methods Some sampling methods are very powerful and enable us to make predictions about large populations from knowledge relatively few members
• Basics of Sampling TheoryPopulation Element Defined target population Sampling unit Sampling frame
• Key Definitions Population – group of things (people) having one or more common characteristics Sample – representative subgroup of the larger population  Used to estimate something about a population (generalize)  Must be similar to population on characteristic being investigated
• Target population As well as defining our topic and formulating our research question, aim or hypothesis, we have also to decide who we want to study This is the target population Most of the time our target population will be too large to include all the potential participants What we do then is select a sample of people from the target population
• Unit of analysis The unit of analysis is the entity that you are analyzing in your study. Any of the following could be a unit of analysis in a study: Individuals Groups (families, households,) Artifacts (books, photos, newspapers) Geographical units (town, census tract, state) Social interactions (dyadic relations, divorces, arrests)
• Unit of Observation Unit of observation is the unit on which data are collected. It is the chosen analysis that determines what the unit of analysis is. For example, a survey may collect data on individuals. That data may include the individuals addresses. Then, in a later analysis, those data can be used to aggregate information to the household level so that the unit of analysis can be the household
• The sampling frame The sampling frame is the list of ultimate sampling unit, which may be people, households, organizations, or other units of analysis. The list of registered students may be the sampling frame for a survey of the students at BPP Telephone directories are often used as sampling frames, for instance, but tend to under-represent the poor (who have fewer or no phones) and the wealthy (who have unlisted numbers).
• Find a sampling frame How would you obtain a sampling frame for the following target populations? Business Students at BPP College New mothers in the Ealing area Households containing 5 or more people in Hammersmith Members of Millwall FC Supporters Club Car dealerships in Stratford
• Size Matters  The larger your sample size, the more sure you can be that their answers truly reflect the population.  This indicates that for a given confidence level (CL), the larger your sample size, the smaller your confidence interval (CI).  However, the relationship is not linear (i.e., doubling the sample size does not halve the confidence interval).
• Sampling ErrorSampling error is any type of bias that is attributable to mistakes in either drawing a sample or determining the sample size
• Factors to Consider in Sample DesignResearch objectives Degree of accuracy Resources Time frame Knowledge of target population Research scope Statistical analysis needs
• Determining Sample Size How many completed questionnaires do we need to have a representative sample? Generally the larger the better, but that takes more time and money. Answer depends on:  How different or dispersed the population is.  Desired level of confidence.  Desired degree of accuracy.
• Factors Affecting Sample Size for Probability Designs Variability of the population characteristic under investigation Level of confidence desired in the estimate Degree of precision desired in estimating the population characteristic
• Common Methods for Determining Sample Size Common Methods:  Budget/time available  Executive decision  Statistical methods  Historical data/guidelines  Sample size calculator http://www.surveysystem.com/sscalc.htm
• Calculating sample size You need to know: The size of the sample frame The confidence interval (CI) The confidence level (CL) http://www.surveysystem.com/sscalc.htm offers a free sample size calculator
• Defining Population of Interest Population of interest is entirely dependent on Management Problem, Research Problems, and Research Design. Some Bases for Defining Population:  Geographic Area  Demographics  Usage/Lifestyle  Awareness
• Types of sampling Sampling methods fall into 2 categories Probability sampling in which elements are selected using a random method All elements have an equal likelihood of selection We are able to estimate error for our sample statistics using probability theory Non-probability sampling does not involve random selection With non-probability samples we may or may not represent the population well, and it will often be hard for us to know how well weve done so.
• Types of Sampling MethodsProbability Sampling Non-Probability Simple random Sampling sampling  Deliberate (quota) Stratified random sampling sampling  Convenience Systematic sampling sampling  Purposive Cluster sampling sampling Multistage sampling
• Random Sampling In a random sample all members of the population have an equal chance of being selected This can be done by lottery Or it can be done using a random number table
• Stratified Sample Sometimes we need to ensure that certain characteristics of our population are represented in the sample (e.g. gender) Stratifying our sample enables us to ensure that they are included We divide the sample into the groups we’re interested in and then sample from the subgroups
• Stratified sample activity You have been appointed to lead a research team assigned with the task of finding the reasons teenagers smoke. The team has decided to conduct a nation-wide survey involving students between 14-16 years of age in secondary schools. Explain how you plan to draw the sample of students using stratified sampling. What subgroups would you include? What further information do you need to draw a representative sample?
• Systematic Sampling  Technique  Use “system” to select sample (e.g., every 5th item in alphabetized list, every 10th name in phone book)  Advantage  Quick, efficient, saves time and energy  Disadvantage  Not entirely bias free; each item does not have equal chance to be selected  System for selecting subjects may introduce systematic error  Cannot generalize beyond pop actually sampled
• Steps in Drawing a Systematic Random Sample 1: Obtain a list of units that contains an acceptable frame of the target population 2: Determine the number of units in the list and the desired sample size 3: Compute the skip interval 4: Determine a random start point 5: Beginning at the start point, select the units by choosing each unit that corresponds to the skip interval
• Cluster Sampling Appropriate when you can‟t obtain a list of the members of the population have little knowledge of pop characteristics Population is scattered over large geographic area Works on the principle that similar elements are often physically clustered together (e.g in neighbourhoods or electoral wards)
• Multistage Cluster Sampling  Stage 1  randomly sample clusters (schools)  Stage 2  randomly sample individuals from the schools selected
• Deliberate (Quota) Sampling  Similar to stratified random sampling  Technique  Quotas set using some characteristic of the population thought to be relevant (e.g. class, gender, ethnicity)  Participants selected non-randomly to meet quotas  Disadvantage  selection bias  Cannot always set quotas for all characteristics important to study
• Volunteer samples  People literally volunteer themselves to participate in the study  Usually in response to an advert, a letter or other inducement  Self-selecting samples can be a problem  Voters in the X factor or magazine polls are typically „self-selecting‟ samples
• Convenience/Opportunitysampling Convenience sampling is sampling which involves the sample being drawn from that part of the population which is close to hand. Convenience sampling is used in research where the researcher is interested in getting an inexpensive approximation of the truth. As the name implies, the sample is selected because they are convenient. Convenience sampling often leads to a biased study since it consists of only available people.
• Convenience Sampling “Take them where you find them” - nonrandom Intact classes, volunteers, survey respondents (low return), a typical group, a typical person Disadvantage: Selection bias Use post hoc analysis to show groups were equal at the start
• Purposive Sampling Purposive sampling (criterion-based sampling)  Establish criteria necessary for being included in study and find sample to meet criteria Solution: Screening  Use random sampling to obtain a representative sample of larger population and then those subjects that are not members of the desired population are screened or filtered out  E.g. want to study smokers but can‟t identify all smokers
• Sample Size  Critical factor is whether sample is representative  Necessary sample size depends on population size  Recommendations:  Use tables from books  30 per group  Descriptive studies – 10-20% of population  No more than 50% of population  Statistical power  Attrition
• When Selecting Subjects … Are subjects with special characteristics necessary for your research? (age, gender, trained/untrained, expert/novice, size, etc.) Can you obtain the necessary permission and cooperation from the subjects? Can you find enough subjects? Interaction among selection of subjects, treatments, and measures is essential for experimental studies.
• Developing a Sampling Plan1. Define the Population of Interest2. Identify a Sampling Frame (if possible)3. Select a Sampling Method4. Determine Sample Size5. Execute the Sampling Plan
• CONFIDENCE LEVELS The confidence level tells you how sure you can be. It is expressed as a percentage and represents how often the true percentage of the population who would pick an answer lies within the confidence interval. The 95% confidence level means you can be 95% certain; the 99% confidence level means you can be 99% certain. Most researchers use the 95% confidence level.
• CONFIDENCE INTERVALS The confidence interval is the plus-or-minus figure usually reported in newspaper or television opinion poll results. For example, if you use a confidence interval of 4 and 47% percent of your sample picks an answer you can be "sure" that if you had asked the question of the entire relevant population between 43% (47-4) and 51% (47+4) would have picked that answer.
• Reporting Subjects State how many subjects were selected Describe how the subjects were selected Discuss whether any subjects were lost during the study and why Explain why the subjects were selected Describe subject characteristics that are pertinent to study – be very specific Identify procedures taken to protect the subjects
• CI & CL When you put the confidence level and the confidence interval together, you can say that you are 95% sure that the true percentage of the population is between 43% and 51%. The wider the confidence interval you are willing to accept, the more certain you can be that the whole population answers would be within that range. For example, if you asked a sample of 1000 people what type of beer they preferred and 60% said Brand A, you can be very certain that between 40% and 80% of all the people in the city actually do prefer that brand, but you cannot be so sure that between 59 and 61% of the people in the city prefer the brand.