Chapter 11.docx

8 Pages
Unlock Document

Management and Organizational Studies
Management and Organizational Studies 4485F/G
Linda Eligh

Chapter 11: Basic Sampling Issues The Concept of Sampling • Sampling is the process of obtaining information from a subset of a larger group • Then using it to estimate the characteristics of the larger group • Must be selected in a manner that is representative Population • Population/universe is a group of elements whom information is needed • Elements can be consumers, companies, stores, houses, cars, trees, dogs, cats • Population questions: Whose opinions are needed to satisfy the objective? What are the characteristics of the target market? Sample Versus Census • Sample: subset of all the members of a population • Ideally representative cross section of total population • Census data obtained from almost all members of the population of interest o Not usually used in market research o Feasible with industrial products that only a small number of customers use • Census does not mean that the results are more accurate • Canada uses the voluntary National Household Survey for census • Budgets and timelines can also create issues with censuses Developing a Sampling Plan • Steps define population, choosing a data collection method, identifying a sampling frame, selecting a sample method, determine sample size, develop operational procedures and executing sampling plan 1. Define the population of interest • Specified in terms of geography, demographic, product or service usage or awareness • Screening process determines who to use • Define who should be excluded o Ask who has family working in market research, advertising or produce area at issue – security question • Tim Horton’s may want to find out why some ppl don’t drink their coffee therefore don’t use ppl who have drank their product 2. Choose a data collection method 3. Identify a sampling frame • A list of the members or elements of the population from which units to be sampled are selected • Specify a procedure for generating such a list • Eg. Telephone book but doesn’t include everyone • Delist their phone numbers are more likely to be renters, live in central city, recently moved, large fam, young kids, lower incomes • Therefore there is major differences between the two groups in terms of purchases, ownership and use of certain products • Only use cell phones: 18-34 have highest usage of only cells at 50% and those who rent • Random digit dialing – specific to industry segments • There is the do not call list but some are exempt o Research firms, charities, political parties, general circulation newspapers and business relations that already exist 4. Select a sampling method • Depends on the objective of the study, finances, time limitations and nature of the problem • Sampling can be divided into probability sampling methods and non probability • Prob samples are selected in which everyone has a known, non zero likelihood of selection and any difference in value is sampling error o Cross section of the population of interest o Sampling errors computed o Results can be projected to total population o More expensive, interview costs and time spent designing sample plan o Systematic, stratified, cluster and simple random • Non probability samples: specific elements from the population have been selected in a non random manner o Don’t chose randomly o Purposeful non randomness occurs when a sampling plan excludes/over represents sertain subsets of the population o Don’t know if it is representative o Sampling error cannot be computed o Cant be generalized o Costs less, quick, representative of a specific population o Snowball, judgment, quota and convenience 5. Determine the sample size • Non prob samples rely on factors like budget, rules of thumb, number of subgroups • Prob samples use formulas to calculate the sample size given the target levels of acceptable error and levels of confidence 6. Develop operational procedures for selecting sample elements • Say what sample method you are using 7. Execute the operational sampling plan • Make sure ppl follow the rules • Ppl can say they don’t want to participate, right to confidentiality • No carte blanche – they cant do whatever they want • Personal Information Protection and Electronics Documents Act o You have to get consent from person o Debriefing after the study Sampling and Non Sampling Errors • Population parameter is a value that defines a true characteristic of a total population – “u” • You can make it from your sample results • You’ll find that in your samples there will be some deviation from the average of the sample before • Two errors: sampling error and non sampling measurement error • Sample mean = true pop mean plus/minus sampling error plus/minus non sampling error • Samplingerror: sample doesn’t perfectly represent the population o Administrative is problems with the execution of the sample/ flaws in the design can avoid by careful attn. to the design and execution o Random sampling error is due to chance and cannot avoid  Little reduction from bigger samples • Non sampling error or measurement error includes all factors other than sampling error o Non response bias  Respondents who respond and don’t respond are different  Ppl choose not to participate  Highest among urban, single person households and households without children  Reduce by encouraging them to respond  Check for differences in late respondents and early respondents  Late respondents tend be similar to non respondents Probability Sampling Methods Simple Random Sampling • Purest form of probability sampling • Probability of selection = sample / population • Sampling frame available can select a simple random sample as follows o Assign number to each element of population then use a table of random numbers to select specific elements to include in the sample • Easy and meets all the requirements of probability sample • Guarantees everyone gets a chance to be picked • Begins with a complete listing of the population (RDD, customer lists, software) Systematic Sampling • Probability sampling in which the entire population is numbered and elements are selected using a skip interval • First number the entire population • Determine a skip interval and selects elements based on this interval • Skip interval = population size / sample size • Random starting point – draw a number • Simpler, less time consuming and less expensive than SRS • Danger that there are hidden patterns within the population Stratified Sampling • Forced to be more representative through simple random sampling of mutually exclusive and exhaustive subsets • Original population divided into two or more sets (male or female) • Random samples taken from the subsets • Don’t tell you how to separate the original subset • Potential for greater
More Less

Related notes for Management and Organizational Studies 4485F/G

Log In


Join OneClass

Access over 10 million pages of study
documents for 1.3 million courses.

Sign up

Join to view


By registering, I agree to the Terms and Privacy Policies
Already have an account?
Just a few more details

So we can recommend you notes for your school.

Reset Password

Please enter below the email address you registered with and we will send you a link to reset your password.

Add your courses

Get notes from the top students in your class.