Test #2 Flashcards
Selection of descriptive research design is based on 3 factors:
- Nature of Problem
- Research questions
- Research objectives
When to use descriptive research design
- When researcher problem is either to describe characteristics of existing market solutions or to evaluate current mktg mix strategies
- If research question include issues s.a. who, what, where, when, and how for target population or mktg strategies
- If task is to identify relationships b.t variables or determine whether differences exist bt groups
Goals of descriptive research survey methods
Provide facts that can be used to:
- Make accurate predictions a/b relationships b/t market factors and behaviors
- Gain insights to understand the relationships & differences
- Verify or validate existing relationships
2 general approaches used to collect data for descriptive research:
- Asking questions
2. Observations
Descriptive research designs often result in the use of _____ to collect ____ data from larger groups through question/answer process
survey research methods
quantitative
With the emergence of_____, observation is being used more often than question/answer in descriptive designs
Scanner data and tracking of Internet behavior
Main goal of quantitative research methods =
to provide facts and estimates from a large, representative sample of respondents
5 advantages of quantitative survey research designs
- Generalizable to target population
- Facilitates advanced statistical analysis
- Accommodates to large number of people
- Concepts and relationships not directly measurable can be studied
- Easy to administered and record answers
3 Disadvantages of quantitative survey research designs
- Questions that accurately measure respondent attitudes and behavior can be challenging to develop
- In-depth data difficult to obtain
- Low response rates can be a problem
Difference between the findings based on the sample and the true values for a population
Sampling errors
Sampling errors are caused by:
Method of sampling used
Size of sample (to reduce error, increase size)
Errors that can occur in survey research design not related to sampling
Nonsampling errors
4 major sources of nonsampling errors
- Project administration errors
- Respondent errors
- Incorrect problem definition
- Measurement/questionnaire design errors
(PRIM)
4 characteristics of nonsampling errors
- Tend to create “systematic variation” or bias in the data
- Unlike random sampling error which can be statistically measured, nonsampling errors cannot be directly measured
- Nonsampling errors are controllable (human mishaps in ether design or survey execution)
- One sampling error leads to nonsampling errors
Consist of both nonresponse error and response error
Respondent errors
Systematic bias that occurs when the final sample differs from planned sample
Nonresponse errors
When respondents have impaired memory or do not respond accurately
Response errors
4 advantages of person-administered surveys
- Feedback
- Adaptability
- Rapport
- Response quality
3 disadvantages of person-administered surveys
- High expense
- Interaction errors
- Possible recording errors
6 advantages of telephone-administered surveys
- Callbacks possible
- Fast
- Interviewers or CATI tech
- Suitable for large samples
- Geographic Flexibility
- Less expensive than face to face methods
4 disadvantages of elephone-administered surveys
- Change in behavior (vm, caller id, mobile)
- Difficult for complex tasks, long surveys, or those using visual aids
- Limited to domestic research
- Perception of telemarketing
4 types of self-administered surveys
- Mail survey
- mail panel
- drop off
- internet
4 Advantages of self administered surveys
- No interviewer response bias
- Anonimity in response
- Respondent control
- Low cost
5 disAdvantages of self administered surveys
- lack of monitoring
- minimal flexibility
- potential response errors
- high nonresponse rates
- slow data acquisitions
3 characteristics in selecting a survey method
- Situational characteristics
- task characteristics
- respondent characteristics
6 situational characteristics when selecting a survey method
- Budget
- completion time frame
- quality requirements
- data completeness
- data generalizability
- data precision
What is data generalizability
projectable to the population represented by sample in a study
degree of exactness of the data in relation to some other possible response
data precision
4 task characteristics when selecting a survey method
- Topic sensitivity
- amount of info
- difficulty of task
- stimuli needed
degree to which a survey ? leads respondent to give a socially acceptable response
topic sensitivity
3 response characteristics when selecting a survey method
- respondent participation
- incidence rate
- diversity
% of the general population that is the subject of the market search
incidence rate
3 things that determine respondent participation
- Knowledge level
- ability to participate
- willingness to participate
casual research designs that can identify cause & effect relationships between variables
experiments
an observable element or attribute of an item or event that can be measured
variable
types of variables in experimental designs
- independent (values manipulated by researcher)
- dependent (measures of effect)
- control (conditions that make the design a true experiment)
- extraneous (uncontrolled, unmeasurable variable that may affect dependent)
extent to which the research design accurately identifies causal relationships
internal validity
extent to which a causal relationship found in a study can be expected to be true for the entire target population
external validity
process of selecting a small # of elements from a larger defined target group of elements such that the info gathered from small group will allow judgments to be made about the larger groups
sampling
To be more confident during a sampling,
you have to increase the margin of error
Identifiable set of elements of interest to the researcher and pertinent to the info problem
population
complete set of elements identified for investigation
defined target population
target population elements available for the selection during the sampling process
sampling unit
list of all eligible sampling units
sampling frame
any type of bias that is attributable to mistakes in either drawing a sample or determining the sample size
sampling error
random sampling error tends to occur because of
chance variations in the selection of sampling units
Sampling errors can be reduced by
increasing the size of the sample
bias that occurs in a research study regardless of whether a sample or census is used, such as bias caused by measurement error, response errors, or coding errors
nonsampling errors
The more ___ a study, the greater the potential for nonsampling errors
extensive
nonsampling error usually related to the ______ whereas sampling errors relate to ___
nonsampling error—accuracy of the data
sampling errors—representativeness of the sample to the defined target population
with this ampling method, each sampling unit in the defined target population has a known probability of being selected for the sample
probability sampling
Sampling method in which the probability of selection of each sampling unit is not known
nonprobability sampling
selection of sampling units is based on:
-judgment of the researcher
may or may not be representative
procedure in which every sampling unit has a known and = chance of being selected
simple random sampling
4 advantages of simple random sampling
- produce unbiased estimates of the population’s characteristics
- easily understood
- results can be generalized to the defined target population with a prespecified margin of error
- valid representation of defined target population
2 disadvantages of simple random sampling
- difficulty of obtaining a complete and accurate listing of the target population elements
- requires that all sampling units be identified
when does simple random sampling work best?
for small populations where accurate lists are available
procedure in which the defined target population is ordered in some way, usually in the form of a customer list, taxpayer roll or membership roster, and selected systematically
systematic random sampling
advantage of systematic random s.
relatively easy way to draw a sample while ensuring randomness
2 disadvantages of systematic random sampling
of sampling units in the target population must be known
possibility of hidden patterns in the list of names that create bias
steps in drawing a systematic random sampling
- obtain a list of units that contains an acceptable frame of the target population
- determine the # of units in the list and the desired sample size
- compute the skip interval
- determine a random starting point
- beginning at that start point, select the units by choosing each unit that corresponds to the skip interval
separation of the target population into different groups (strata) and the selection of samples from each stratum
stratified random sampling
3 steps in drawing a stratified random sample
- divide the target population into homogeneous subgroups or strata
- draw random samples from each stratum
- combine the samples from each stratum into a single sample of the target population
2 methods of stratified random sampling
- proportionately stratified sample
2. disproportionately…
in this s. method, each stratum is dependent on its size relative to the population
proportionately stratified sample
in this s. method, each stratum is independent on its size relative to the population
disproportionately stratified sample
3 advantages of stratified rs
- assurance of representativeness in the sample
- opportunity to study each stratum and make comparison b/t strata
- ability to make estimates for the target population with the expectation of greater precision and less error
2 disadvantages of stratified rs
- determining the basis for stratifying
- info relevant to the required stratification factors might not be readily available therefore forcing the researcher to use less desirable criteria as the factors for stratifying the target population
method in which sampling units are divided into mutually exclusive and collectively exhaustive subpopulations
cluster sampling
form of cluster sampling in which the clusters are formed by geo designations
area sampling
2 advantages of cluster sampling
cost effective
easy to implement
2 dis. of cluster sampling
- clusters are often homogeneous
2. appropriateness of the designated cluster factor used to identify the sampling units within clusters
4 types of probability sampling
- simple random s.
- systematic rs
- stratified rs
- cluster sampling
Sampling method in which samples are drawn at the convenience of researcher
convenience sampling
advantage of convenience sampling
enables a large 3 of respondents to be interviewed in a relatively short time
3 disad of convenience sampling
- constructs mights be unreliable if used to study a broader target population
- data are not generalizable to the defined target population
- representativeness of the sample cannot be measured bc sampling error estimates cannot be calc
sampling method in which participants are selected according to an experienced individual’s belief that they will meet the requirements of the study
judgment sampling
“purposive sampling
advantage of judgment sampling
if the judgment of the researcher is correct, the sample generated by judgment sampling will be better than one generated by convenience sampling
sampling in which participants are selected acc to pre-specified quotas regarding demographics, attitudes, behaviors or some other criteria
quota sampling
3 advantages of quota sampling
- sample generated contains specific subgroups in the proportions desired by researchers
- appropriate subgroups are identified and included in the survey
- reduce selection bias by field workers
2 disad of quota s.
- success of study dependent on subjective decisions made by researchers
- representativeness of sample cannot be measured
sampling in which set of respondents are chosen and they help researcher identify additional ppl to be included in study
snowball sampling
Snowball/referral sampling is used when:
- defined target population is small and unique
2. compiling a complete list of sampling units is very difficult
advantages of snowball s.
- good way to reach members of small, hard-to-reach, uniquely defined target population
- most useful in qualitative research
2 disad of snowball s.
- allows bias to enter study
2. ability to generalize the results to members of targett population is limited
4 types of nonprobability sampling methods
- Convenience sampling
- judgment s.
- Quota s.
- Snowball s.
7 factors to consider in sample design
- Research objectives
- degree of accuracy
- resources
- time frame
- knowledge of target population
- research scope
- statistical analysis needs
3 factors affecting sample size for probability designs:
- Level of confidence desired in estimate
- variability of the population characteristic under investigation
- Degree of precision desired in estimating the population characteristic
certainty that the true value of what we are estimating falls within the precision range we have selected
confidence
acceptable amount of error in sample estimate
precision
For a particular sample size, there is a trade off between degree of ______ and degree of _____
Degree of confidence & degree of precision
When the defined target population size in a consumer study is 500 elements or less, the research should consider taking a ____
census of the population rather than a sample
Blueprint or framework needed to ensure that the data collected are represented of the defined target population
sampling plan
7 steps to developing a sampling plan
- define target population
- select the data collection method
- identify the sampling frame(s) needed
- select the appropriate sampling method
- determine necessary sample sizes and overall contact rates
- create plan for selecting units
- execute the operational plan
process of assigning intensity (amount) to the info about constructs, concepts and objects
measurement
Measurement of a sample size consists of 2 tasks
- Construct development
2. Scale measurement
Hypothetical variable made up of a set of component responses or behaviors that are thought to be related
construct
constructs are made up of a combination of ____
several related indicator variable that together define the concept being measured
integrative process in which resarchers determine what specific data should be collected for solving the defined research problem
construct development
process of assigning descriptors to represent the range of possible responses to a question about a particular object or construct
scale measurement
combination of labels in scale measurement
scale descriptors
designated degrees of intensity assigned to the responses in a given questioning or observation method
scale points
all scale measurements can eb classified as 1 of 4 basic levels
- Nominal scale
- ordinal scale
- interval scale
- ration scale
this scale measurement focuses on only requiring respondent to provide some type of descriptor as the raw response (married/single)
nominal scale
type of scale measurement that allows a respondent to express “relative’ magnitude between the raw responses to a ?
ordinal scale
this scale measurement demonstrates the absolute difference between each scale point (strongly agree/strongly disagree)
interval scale
scale measurement that allows for the identification of absolute differences between each scale point and absolute comparisons between raw responses
ratio scale
Central Tendency & dispersion
central tendency –> mode, mean, median
Dispersion –> frequency distribution, range, SD
this scale measurement demonstrates the absolute difference between each scale point (strongly agree/strongly disagree)
interval scale
scale measurement that allows for the identification of absolute differences between each scale point and absolute comparisons between raw responses
ratio scale
Central Tendency & dispersion
central tendency –> mode, mean, median
Dispersion –> frequency distribution, range, SD
If a nominal scale is used, analysis of the raw data can only be done using ___ and ____
modes & frequency distributions
If ordinal scale is used, analysis of raw data can be done using ___ and ____
medians & ranges
plus modes and frequency distribution
If interval/ratio scales are used, analysis of raw data can be done through the use of ___ and ____
sample means & estimated SD (s.a. sample statistic)
Construct/scale development process
- identify and define construct
- Create initial pool of attribute statements
- Access and select reduced sets of items
- Design scales and pretest
- Complete statistical analysis
- Refine and purify scales
- Complete final scale evaluation
ordinal scale format that asks respondents to indicate the extent to which they agree or disagree with a series of mental or behavioral belief statements about a given object
Likert Scale
Unique bipolar ordinal scale format that captures a person’s attitudes and/or feelings about a given object
Semantic differential scale
with the semantic differential scale, only the ___ of the scale are labeled
endpoints
Disadvantage of Semantic Differential Scale
inappropriate narrative expressions of the scale descriptors
special type of rating scale designed to capture the likelihood that people will demonstrate some type of predictable behavior intent toward purchasing an object or service in a future time frame
behavioral intentional scale
This scale is a good predictor of consumers’ choices of frequently purchased and durable consumer goods
Behavioral intentional scale
scale that requires a judgment without reference to another object, person or concept
noncomparative rating scale
Scale that requires a judgment comparing one object person or concept against another on the scale
Comparative rating scale
This scale uses a scale point format that presents respondent with some type of graphic continuum as the set of possible raw responses to a given question
Graphic rating scale
4 Scale measurement issues
- single-item or multiple-item scales
- clear wording
- Screening questions
- skipping questions
scale format that collects data about only one attribute of an object or construct
single item scale
formal framework consisting of a set of questions/scales designed to generate primary data
questionnaire
This involves using a process that takes established sets of question/scale measurements and formats them into a complete instrument
questionnaire construction
7 Steps in questionnaire designs
- Confirm research objectives
- select appropriate data collection method
- develop ?s and scaling
- determine layout and evaluate questionnaire
- obtain initial client approval
- pretest, revise, and finalize questionnaire
- implement survey
open-ended format where respondent replies in their own words
unstructured questions
close ended format where respondents respond from a set of possible responses
structured questions
3 qualities of bad questions
- unanswerable
- leading or loaded
- double barreled questions (asking two things at once)
Guidelines for evaluating the adequacy of questions
- Use simple words (no technical)
- avoid qualifying phrases
- ensure response categories are mutually exclusive
- ensure question and scale statements are meaningful to the respondents
- avoid arranging response categories in a manner that may bias responses
- do not double-barrel questions
Considerations in questionnaire design
- Confirm the research objectives before starting
- determine data reqs to complete objectives
- include general description of study in intro
- ensure instructions are clear
- arrange ?s in a logical order
- begin with simple questions and then progress to more difficult ones
- ask personal/sensitive questions at the end
- avoid asking questions with diff measurement formats
- end with thank you
separate written communication to prospective respondent designed to enhance that person’s willingness to participate
cover letter
Guidelines for developing cover letter
-personalize
identify sponsoring org
-state purposely and importance clearly
-promise anonymity or confidentiality
-clarify the general time frame of doing study
-reinforce importance of participation
-acknowledge reasons for non participation
-time requirements and any compensation
-explain completion data and where and how to return
-advanced thank you
Every set of data collected needs some summary info developed containing: (3)
- Central tendency and dispersion
- Relationship of sample data
- Hypothesis testing
Measure of central tendency that describes distribution best for nominal, ordinal, interval/ratio data
nominal –> mode
ordinal –> median
interval/ratio –> mean
describes how close to the mean or other measure of central tendency, the rest of values fall
measures of dispersion
distance b.t smallest and largest value in setq
range
avg distance of the distribution values form mean
SD
two or more groups of responses that are tested as though they may come from different populations
independent samples
two or more groups of responses that originated from sample population
related samples
used for 30+ ppl in sample
z-test (or unknown #)
univariate test of significance is appropriate for this kind of data
interval or ratio
test to compare charac of two groups or two groups
bivariate statistical test
3 things bivariate statistical test consists of
- cross-tabulation with chi-square
- t-test to compare two means
- analysis of variance (ANOVA) to comapre three or more means
enables researcher to test for statistical significance between the frequency distributions of 2 or more nominally scaled variables in a cross tabulation table to determine if there is any association between the variables
chi-square analysis
Comparing means requires this type of data
ratio/interval
difference between the means divided by the variability of random means
t-test
ration of difference between the two sample means and the std error
t-test
T-test tries to determine if the difference between the two sample means ____
occurred by chance
statistical technique that determines if three or more means are statistically different form each other
Analysis of valiance (ANOVA)
In order to perform ANOVA, the dependent variable must be ______ and the independent variable must be ____
dependent variable = measurable (interval/ratio scaled)
Independent variable = categorical
What is one-way ANOVA?
only one independent variable
this test is used to statistically evaluate the differences between the ggroup means in ANOVA
f-test
in ANOVA, total variance in a set of responses can be separated into ___ group and ___ group variance
between group and within group
the larger the diff in the variance b/t groups, ____ (ANOVA)
the larger the f-ratio
ANOVA does not tell us____
where the significant lies
process used to develop maps showing the perceptions of respondents
perceptual mapping
this test is used to statistically evaluate the differences between the ggroup means in ANOVA
f-test
in ANOVA, total variance in a set of responses can be separated into ___ group and ___ group variance
between group and within group
the larger the diff in the variance b/t groups, ____ (ANOVA)
the larger the f-ratio
ANOVA does not tell us____
where the significant lies
Quantitative data uses numbers and stats to summarize _______
Demographics
Attitudes
Behaviors
process used to develop maps showing the perceptions of respondents
perceptual mapping