Scribd
Upload
44 views13 pages

BUSI 5801 Mod 3 Sources Slides

Uploaded by

GouravRaj
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
44 views13 pages

BUSI 5801 Mod 3 Sources Slides

Uploaded by

GouravRaj
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 13

3.

Sources and Samples


BUSI 5801
Statistics for Managers
Online
Module 3 Learning Objectives: Sources
and Samples
At the end of this module students will be able to:
• Recognize the basics of sampling concepts and terminology
(CO 1)
• Identify what sample information is needed to interpret data
(CO 1,4)
• Determine possible sources of sampling limitations or bias
(CO 4)
Sample versus Population
• A population represents all of the elements or units (people,
companies, countries, etc.) in which you are interested
• A sample generally represents the subset of the population from
whom/which you have collected data
• The sample = the who of the data
• We usually employ samples because collecting data from entire
populations is costly, time consuming or impossible (called a
census)
Sample versus Population
• We use data from sample to infer insights into the entire
population – inference is imprecise
• We sacrifice precision (using a sample) for practicality (using
entire population)

• Parameter: Describes a population (e.g., population mean)


• Statistic: Describes a sample (e.g., sample mean)
Sampling Frame
• A sampling frame is the specific set of units/elements in
your population that you want to study.
• Ideally, the sampling frame is the same as the population
• But it may be more limited based on practical reasons (all
residents of a city versus list of everyone with an address)
• The sampling frame clarifies from where you will draw your
sample
Target Sample
• A target sample is the subset of the sampling frame from
whom/which you intend to collect data
• There is an infinite number of possible subsets within a
sampling frame
– both with respect to who and how many
• The particular subset that makes up a sample has
implications for the meaning of the analysis
• A biased sample contains an over- or under-representation
of some characteristics of the population and may lead to
biased results
Randomization
• A randomized sample is most likely to be representative of
the population across all characteristics
• A randomized sample minimizes bias – but does not
eliminate it
• In a randomized sample:
– All elements of the population have an equal probability of being
included in the sample
– The elements of the sample are chosen randomly
• Note: no two randomized samples are likely to be exactly
the same
Other “Randomized” Sample Designs
• More complicated sample designs may save time or money
or avert sampling problems
• Stratified Sampling, Cluster Sampling, Systematic Sampling,
Multistage Sampling
• These require dividing or ordering the population into
groups based on some characteristics and sampling
randomly from the groups/lists
• Because these methods require choosing the
characteristics upon which to divide or order the population,
the benefits of randomization are sacrificed
Sample Size
• The bigger the sample, the better it will represent the
population
• The size of the sample affects what can be concluded from
the analyses
• The size of the sample usually depends on:
– What we want to do with the data
– Practical limitations (remember – this is why were are not studying a
whole population)
Sampling

Population Sampling Target


Frame Sample
Sample
Sampling Biases
• Broad invitations to participate (voluntary sampling – low, biased
response rates)
• Convenience sampling (including those that are convenient)
• Response biases (providing the “right” answer)
• Non-response bias (portion of sample cannot or will not respond)
• Inattention (long surveys may lead to inattention and low quality
results)
• Ineffective, confusing or leading questions
Parameters and Statistics
• A representative characteristic of a population is a
parameter
• The same characteristic of a sample is a statistic
• We used sample statistics to estimate the parameters of
populations
• Examples: means, standard deviations, correlations, etc.
Recap: 3. Sources and Samples
• Sample data tells us about the sample
• But we often use samples to make inferences about populations
of interest
• The subset that you target for your sample can bias your results
– randomization is best
• The subset of your target sample that ends up in your final
sample will affect your results
• The approach to solicit participation in your study will affect your
results
• The methods used to collect data (e.g. surveys) will affect your
results

You might also like