When measuring every item in a population is impossible, inconvenient or too expensive, we intuitively take a sample. Sampling is a central aspect of marketing research and does much to determine how realistic marketing results will be.

We draw conclusion about an entire population by taking measurements from only a portion of all population elements. We estimate characteristics of population from sample estimates.

Why we choose samples? Convenience, budget constrains, it is accurate and reliable! (properly selected samples give results that are reasonably accurate, the larger the sample size, allows one to draw more accurate conclusion about the population.

A sample is a subset or some part of a larger population.

A population (universe) is any complete group (i.e., people, sales territories, stores, etc.) sharing some common set of characteristics.The term population element refers to an individual member of the population.

A census is an investigation of all the individual elements making up the population—a total enumeration rather than a sample.

Determining Target market population

sampling fram - a list of elements from which a sample may be drawn; also called working population.

Sampling Frame Error

…occurs when certain sample elements are not listed or are not accurately represented in a sampling frame.

A list of elements from which the sample may be drawn

Working population

Mailing lists - data base marketers- phone directories,university email directory, student email directory could not be expected to perfectly represent the student population; perfect representation is not always needed or possible.

Sampling services: companies specialized in providing lists or data bases including names, addresses, phone numbers etc. They compile these lists from subscriptions to  professional journals, credit card applications, etc.

Online panels: list of respondents who have agreed to participate in marketing research along with the email contact information for these people. Ex: qualtrics. online panel= sampling frame

Random sampling error: refers only to statistical fluctuations that occur because of chance variations in the elements selected for the sample.

Random sampling error is a function of sample size.

We previously talked about two major sources of error: random sampling error and systematic error;

Random sampling error: the difference between the sample result and the result for the population. Random sampling error occurs because of  chance variation in the selection of sampling unit. Random sampling error is a function of sample size.

Systematic error: refers to the nature of the study’s design and execution. Ex: some people such as old ones may not be available in online panels….

Exhibit shows that sampling frame error and nonresponse error are nonsampling errors related to design.the total population is represented by the area of the largest square. Sampling frame errors eliminate some potential respondents. Random sampling error then comes in play. Additional error will occur if individuals refuse to respond or can not be contacted.Overall the sample may over represent or under represent specific portion of the population.

Probability

Probability Sampling

Random refers to the procedure for selecting the sample, it does not describe the data in the sample. Randomness should not be thought of as an unplanned or unscientific. It is the basis of all probability sampling methods.

Each population member, and each possible sample, has equal probability of being selected.

Equal chance of selection. Ex: drawing named from a hat/ rolling a dice/ using cards. This process is simple and only one stage of sample selection is required. When we have large populations, sample selection can be based on tabled random numbers or computer-generated random numbers.

How to pick random numbers: assign numbers to each member of the population…Using random number table is a way for that (conventional). Put your finger somewhere to start and then pick a direction to go..computer generated random numbers is a better way.

Sometimes it is useful to divide the population into some subgroups/strata (if researcher is interested to investigate some differences between strata). It can be more efficient than simple random sampling and smaller standard error may result from this method because the groups will be adequately represented when strata are combined.

Ex: male and females samples; seniors and juniors, major, based on the type of phone.

Proportionate Stratified Sample

Number of objects/sampling units

chosen from each group is

proportional to number in population

Disproportionate Stratified Sampling

Sample size in each group is not

proportional to the respective group

sizes

Used when multiple groups are

compared and respective group sizes

are small

When we want to ensure an adequate number of sampling units in every stratum. The sample size is dictated by analytical considerations.

Rule of thumb: Stratum size increases for strata of larger sizes with the greatest relative variability. (we know that sample size should increase when we have higher variability in population in general, here strata.

The primary sampling unit is no longer the individual element in the population but a larger cluster of elements located in proximity to one another (ex: cities).

That’s why it is probability sampling.

It is useful when list of the sample population does not exist. Ex: employees, by organizations.

Very cost effective

Nonprobability Sampling

…costs and trouble of developing sampling frame are eliminated

Results can contain hidden biases and uncertainties

Convenience

Used to obtain information quickly

and inexpensively

Sampling by obtaining people or units that are conveniently available. Ex: interview customers at a shopping center/ asking your friends to fill your questionnaire/ visitors of a website (google) from social discussion pages…

Costs and trouble of developing sampling frame are eliminated

Results can contain hidden biases and uncertainties

Projecting the results beyond the specific sample is inappropriate. Best used for exploratory research when additional research will subsequently be conducted by probability sampling. Ex: student samples that professors use, can not be generalized to a larger population. (internal versus external validity) internal validity: effects hypothesized hold true under maximum control of outside effects.

                       Judgmental

"Expert" uses judgment to identify

representative samples

An experienced individual selects the sample based on his/her judgment about some appropriate characteristics required of the sample member.

Quota 

Minimum number from each

specified subgroup in the

population

The purpose of quota sampling is to ensure that the various subgroups in a population are represented. The difference with stratified sampling is that here in quota researcher has a quota to achieve. 35 interviews with juniors, 22 with seniors, etc… interviewer is responsible to find enough people to meet the quota.

It introduces bias…we will see for example all people under age 25 are were college educated because interviewer picked them…Quota sample tend to include people who are easily found, willing to be interviewed, and middle class.

Advantage: speed to data collection, lower costs and convenience. 

Snowball

- Form of judgmental sampling

- Appropriate when reaching small,

specialized populations

- Each respondent, after being interviewed, is asked to  

identify one or more others in the

appropriate group

Define: using probability methods for an initial selection of respondents and then obtaining additional respondents through information provided by the initial respondent.

Advantage: reduced costs

Disadvantage: bias because a person suggested by someone already in the sample has a higher probability of being similar to the first person.

Determining the Sample Size

Ad Hoc Methods

Where:

 z is the standardized value at a specified confidence level
s is the sample standard deviation
E is sampling error