MKTG2010 Review

• Management decision problem.
• Action oriented
• Identifies what actions should be taken
• Focus on symptoms
• eg: Should product X be introduced into the market

• Marketing research problem (will answer the MDP)
• Information oriented
• What information is needed
• How can it be obtained in the most feasible way?
• Focus on underlying cause
• eg: Assess possible market share and profit for product X

• Marketing Research Objective (will guide the research)
• Often begin with such words as:
• to determine
• to compare
• to rank
• to identify

• Cognitive (thoughts)
• Connative (behaviours)
• Affective (feelings

• Exploratory
• (Conclusive) Descriptive
• (Conclusive) Causal

• Discovery of ideas and insights
• Flexible, versatile and often the front end of the total research design.
• To understand...develop hypotheses...why...?
• Findings can be used in their own right or lead to conclusive research .
• Used for questionnaire design.
• Methods include: expert surveys, pilot surveys, secondary data and qualitative research.

• Use in situations when the background problem is unclear Researchers may conduct exploratory research to:
• Diagnose a situation
• Helps set research priorities
• Helps orientate management
• Screen alternatives
• Select products for trial
• Concept testing (procedure that tests a stimulus as a proxy for an idea Discover new ideas )
• Generate new ideas for products, advertising copy.
• Discover consumer needs

• To measure the state...what is...?
• Describes characteristics or functions of a population or phenomenon.
• Asks who, what, when where, why and how.
• Can also include numerical analysis
• Marked by the formulation of specific hypothesis
• Methods include: secondary data, surveys, panels, observation and other data.

• An experiment used to obtain evidence of cause-and-effect (causal) relationships.
• Involves manipulation of one or more independent variable.
• Causal research is used for the following purposes:
• To understand which variables are the cause (independent variables) and which variables are the effect (dependent variables) of a phenomenon.
• To determine the nature of the relationship between the causal variables and the effect to predict.
• Causality can NEVER be proved (demonstrated conclusively). It can only be inferred that a cause-and-effect relationship exists

Personal face-to-face in-home interviews or Central location personal interviews

• Probably highest response rate
• Use of complex questions
• Probing of open-ended questions
• Clarification of ambiguous questions
• Interview time can be long

• Generally narrow distribution
• Expensive method of administration
• Costly to revisit “not-at-homes”
• Slow method of administration
• Subject to interviewer bias

• Internet
• Computer administered interviewing (CAI)
• E-mail

• Speed of data collection
• Worldwide distribution possible
• Generally low cost
• Respondents anonymous or known
• Can be used for sensitive topics
• Visuals can be incorporated
• Not subject to interviewer bias

• Cannot be too long nor too complex
• Cannot explain ambiguous questions
• No probing with open-ended
• Response bias
• Difficult to establish representative sampling frame

• Stratified
• Systematic
• Simple Random
• Cluster

• Each element has a known and equal chance of selection
• Every element is selected independently of every other element
• Sampling frame is compiled whereby each element is assigned a unique ID
• Random numbers are generated to determine which elements are included in the sample. Eg: 24 people in sampling frame, randomly choose 6 and approach them to sample.

Select random starting point and selecting every nth element in succession Assume population elements are ordered

• Population is split into mutually exclusive and collectively exhaustive sub-populations.
• Then Elements are selected from each stratum by a random procedure (typically SRS)

Target population is divided into mutually exclusive and collectively exhaustive clusters. A random sample of clusters are randomly selected.

• Response error:
• Researcher errors
• Interviewer errors
• Respondent erros
• Non response error

• Researcher errors: poor planning
• Interviewer errors: occur when collecting the data
• Respondent selection: interviewer lacks the skills or motivation to seek out appropriate respondents
• Questioning error: may phrase incorrectly or use wrong tone.
• Recording error: not recording verbatim, ticking the wrong boxes
• Cheating error: interviewer may cheat the complete the required number of surveys.
• Respondent errors:
• Inability error
• Unwillingness error

• Potential respondents included in the sample do not respond
• Sample estimates will be biased if non-respondents differ from respondents on key characteristics
• Refusals
• Non-at-homes

• Nominal
• Ordinal
• Scale - interval
• Scale - ratio

• eg Gender, store types
• Numbers serve as a label identifying objects
• Objects are viewed as equivalent
• Number does not reflect the amount of a characteristic a object possesses
• Statistical analysis based on frequency counts are permissible
• Not meaningful to compute an average

• eg Quality rankings
• Numbers are assigned to objects to indicate the relative extent to which the object possess a characteristi
• c Indicates relative position, not magnitude of the difference between objects

• Eg: attitudes and opinions
• Equal distances on the scale represent equal values in the characteristic being measured
• Allows comparison between objects
• Difference between 1 and 2 is the same as 2 and 3 etc.
• Location of the 0 point is not fixed

• Eg length, weight, age
• Equal distances on the scale represent equal values in the characteristic being measured
• Allows comparison between objects: Difference between 1 and 2 is the same as 2 and 3 etc.
• Absolute 0 point

The systematic and objective process of generating information to aid in making marketing decisions. Outline the marketing research process.

• Relatively unstructured
• Primarily exploratory design based on small samples (can also be used for descriptive research
• Intended to provide insight and understanding

• Seeks to quantify data
• Typically applies some form of statistical analysis
• Testing or confirming theory

• Structured or unstructured: Researcher specifies what is to be observed and how the measurements are to be recorded compared to when the observer monitors all aspects of the phenomenon that appear relevant to the problem at hand
• Undisguised or disguised: Subjects are aware that they are being observed compared to when they are unaware. Eg mystery shopping. In a natural (where their behavior normally takes place) or contrived setting (in an environment that has been specially designed for recording their behaviour)
• Via human (Individuals are trained to systematically observe a phenomenon and to record on the observational form the specific events that take place) or via mechanical administration such as a peoplemetre where a mechanical device observes a phenomenon and records the events that take place

• Observation: the situation is watched and the relevant facts, actions or behaviour are recorded.
• Natural: Subjects are observed in the environment where the behavior normally takes place.
• Contrived: Subjects are observed in an environment that has been specially designed for recording their behaviour
• Personal: Individuals are trained to systematically observe a phenomenon and to record on the observational form the specific events that take place
• Mechanical: A mechanical device observes a phenomenon and records the events that take place (e.g., peoplemeters)
• Audit: Data is collected by examining physical records or performing

The primary factor or controllable variable. Anything that can vary (intensity or value), includes factors.

Primary effectors of the outcome. What the researcher is trying to understand. Will change as a result of the independent variables.

Closely related to independent variable. Consequences of the cause and effect relationship 

Alter the manner and intensity of the relationship between the independent, intervening and dependent variables

• Extraneous variables: All variables other than the independent variables that affect the response of the test units. Can confound the dependent variable measures in a way that weakens or invalidates the results of the experiment.
• Classified as:
• History (H): Events which are external to the experiment but occur at the same time as the experiment
• Maturation (MA): Changes to test units not caused by the impact of independent variables, but rather the passage of time
• Selection Bias (SB): Improper assignment of test units to treatment conditions. When assignment of test units results in treatment groups that differ on the dependent variable before exposure to treatment
• Mortality (MO): Loss of test units while the experiment is in progress

• Convenience
• Quota
• Snowball
• Judgement

May yield good estimates of population characteristics. Estimates not statistically projectable to the population

• A sample of convenient elements (right place, right time).
• Not appropriate for research projects involving population inferences.
• Advantages:
• Least expensive and time consuming.
• Sampling units are accessible, easy to measure and co-operative
• Disadvantages:
• Selection bias
• Not representative
• Cannot generalise to a population

• Two-stage restricted judgement sampling
• Develop control categories (or quotas) of population elements. (Eg: Sex or Age)
• Sample elements are selected based on convenience or judgement

• Initial group of respondents selected (often at random)
• Initial respondents asked to identify others who belong to the target population of interest. And so on....
• Estimate characteristics that are rare in the population
• Locate heavy-users of product categories or brands

• 1. Establish hypotheses
• 2. Your hypothesis should tell you whether you want to test: The shape of the distribution (univariate) A difference (bi-variate) An association (bi-variate)
• 3. Check the shape of your scale data
• 4. Use the appropriate test

• Defining the research problem
• Planning the research design
• Planning the sample collecting the data
• Analysing the data
• Formulating conclusions and preparing the report

• (these conditions are necessary but not sufficient to demonstrate causality).
• Concomitant variation: The extent to which a cause, X, and an effect, Y, occur together or vary together in the way predicted by the hypothesis under construction.
• Time order of occurrence of variables: The causing event must occur either before or simultaneously with the effect, it can not occur afterwards
• Elimination of other possible causal factors: The factor (or variable) being investigated should be the only possible causal explanation. Controls could be: pricing, advertising, level of distribution or product quality. We must satisfy them but we can’t say we 'proved'

7-point rating scale with end-points associated with bi-polar labels cold – warm

• Symmetric distribution (bell-shaped): Values on either side of the centre of the distribution are the same
• Skewed distribution: Deviations from the mean are larger on one side than the other
• Skewness value closer to 0 than [1] suggests a normal distribution.
• Negative numbers mean negatively skewed (skewed to the left)
• Positive numbers means positively skewed (skewed right)

• Parametric Tests:
• Variables are measured on at least an interval scale
• Variables are normally distributed
• Non-parametric Tests: 
• Variables are measured on a nominal or ordinal scale
• Also for when interval/ratio data fails to satisfy parametric test requirements (i.e., normality)

• [.1]-[.3] = weak
• [.3] - [.7] = moderate
• [.7] - [+1] = strong