Lecture 1: Introduction / Overview of Consumer Behavior
Introduction
Course goal 1: understand consumers
- Acquire knowledge on relevant theories and frameworks
- Understand how to obtain consumers insights & what is their role and relevance in
marketing practice
- Learn to apply these to address issues in marketing and public policy
Course goal 2: develop academic skills
- Develop analytical, logical, critical, nuanced thinking
- Read and understand academic research on marketing and consumers
- Understand the experimental research method
Overview of Consumer Behavior
Consumer behavior: the totality of decisions about
the consumptions of an offering by decision-making
units over time
How to get wise about consumers?
- Watch each decision-making stage (5 W’s)
o Who is the consumers?
o What is the consumer doing/looking
for/considering?
o When does the consumers make the purchasing decision, and what’s
going on before, during, and after?
o Where are consumers making decisions? (can vary by stage!)
o Why are they doing this?
- Asses the internal factors in the stage
o Human psychology: beliefs, attitudes, values, motives (drivers of
behavior), needs, wants, emotions (happiness, regret, anger, fear,
pride gratitude, shame, guilt), judgement and decision making,
information processing
- Consider the social factors
o Peers, other consumers, society: self-presentation, social
comparison, social norms, social motives, social influence tactics
- What aspects of the environment influence consumers? (external influences)
o Managerial, market, environment:
§ Market influences, economic factors: resource scarcity, economic cycles
§ Social and cultural factors: political climate, trends, events
§ Context eFects: time of the day, physical environment
§ Choice architecture (store lay-out and other factors)
,Lecture 2: Consumer Research / Studying the CSR Halo
Consumer Research
Surveys & focus groups are often inaccurate because:
- People don’t know what caused their behavior
- Memory is inaccurate
- People are bad at predicting future behavior
- People often don’t tell you what they know, because: social desirability, demand
characteristics, privacy/taboos
Scientific method: observe à question à hypothesis à method à result/conclusion
- Studying actual behavior overcomes most of this, that’s why we use experiments and
field studies
Correlational research strategy:
measures two (or more) variables for
each unit (consumer, firm, brand, etc.)
- E.g., relationship between ad
expenditures and sales
- Correlation is not the same as
causation
- Spurious correlation: two or more variables are
associated but not causally related to each other
o E.g., Is the number of crimes committed caused by
the number of ice creams eaten?
Basic elements of an experimental study:
- Manipulation: researcher manipulates one variable by
changing its value to create a set of two or more
treatment conditions
- Measurement: a second variable is measured for a group of participants resulting in
a set of scores in each treatment condition
- Comparison: the scores in one treatment condition are compared with the scores in
another treatment condition
- Control: all other variables are controlled to be sure that they do not influence the
two variable being examined
Terminology for experimental research:
- Independent variable (IV): the variable that is manipulated by the researcher by
creating a set of treatment conditions
o Condition: a situation or environment characterized by one specific value of the
manipulated variable (IV)
§ Levels: di\erent values of the IV, selected to create and define the treatment
conditions
- Dependent variable (DV): the variable that is observed for changes to assess the
e\ects or manipulating the IV
,- Extraneous variables: all other variables in the study other than the IVs and DVs
A/B tests: compares two versions of a (product/website/app/marketing message/etc.)
to see which one is the better performer
- E.g., Which color scheme will increase click rate?
o IV: color schema
o DV: click rate
o 2 conditions: blue & green
Within subjects design: single group of
participants
- Manipulates the IV by creating di\erent
conditions or treatments
- Measures the DV after each treatment
- Uses a single group of participants, who are exposed to each condition
- Advantage: no variation between subjects on other characteristics, so that your
statistical power is high
- Disadvantage: exposures can influence each other (this may be tackled by a
balanced or randomized presentation order)
Between subjects design: multiple groups of participants
- Manipulates the IV by creating di\erence conditions or treatments
- Measures the DV after each treatment
- Uses multiple groups of participants, each group is exposed to one condition
- Advantage: exposures cannot influence each other
- Disadvantage: participants may vary on other characteristics (this is tackled by
random assignment, which works better when groups are larger, so any
coincidental di\erences cancel out)
Minimizing variance within treatments
- Less variance = more precise and better interpretable results!
- Standardize procedures and treatment settings
- Random assignment and matching
o Limits confounding from individual di\erences
o Sample size: larger samples help reduce variance
- Minimize individual di\erences
o Create equivalent groups
o Reduce variance within groups
o You can use covariates to correct for remaining variance
- Random assignment allows for causal claims
- Random sampling creates representativeness
Analyzing results: statistical significance
- Develop a hypothesis about the e\ect of the treatment
- Conduct a test to examine if di\erence between groups (= the treatment e\ect) is
statistically significant
- What does a significant result mean?
, o The di\erence is large enough and consistent enough to rule out chance as a
plausible explanation (variance between groups is clearly larger than variance
within groups) à we can be reasonably sure (e.g., 95%) that the di\erence is
caused by the treatment
Applications and statistical analyses of between-subjects designs:
- Single-factor design with two groups
o Simples version of between-subjects experimental design
§ Compares the mean scores on DV of two groups of participants
§ Researcher manipulates one IV with only two levels
o Independent-measures t-test (or a single-factor ANOVA with two groups) tests if
the means are significantly di\erent
- Single-factor design with multiple groups
o Compares di\erent instances of a qualitative variable (e.g., color) or di\erent
levels of a quantitative variable (e.g., price)
o Overall e\ect is tested by a single-factor (“one-way”) ANOVA with multiple
groups
o Follow-up comparisons are needed to test di\erences
between specific pairs of groups/levels
Mediation: why does the effect occur? à mediation can be
addressed statistically by measuring an additional carriable
that captures the underlying psychological mechanism that
‘explains’ the effect
- NOTE: mediation is only correlational!
Moderation: what can change the effect? à academic
research uses multi (two-) factor designs to examine
moderation, moderators can be manipulated or measured
- From whom or in which situations does this e\ect occur?
Types of moderation
- No moderation: graph 1
o Only a main e\ect of IV (left panel)
o Main e\ects of both variables (right panel)
- Positive moderation: graph 2
o Strengthens the e\ect or makes it appear ‘necessary
condition’
- Negative moderation: graph 3
o Weakens the e\ect or makes it disappear ‘boundary
condition’
- Crossover moderation: graph 4
o Changes the direction of the e\ect: without main e\ects
(left panel), with a main e\ect of moderator (right panel)
Studying the CSR Halo
Introduction
Course goal 1: understand consumers
- Acquire knowledge on relevant theories and frameworks
- Understand how to obtain consumers insights & what is their role and relevance in
marketing practice
- Learn to apply these to address issues in marketing and public policy
Course goal 2: develop academic skills
- Develop analytical, logical, critical, nuanced thinking
- Read and understand academic research on marketing and consumers
- Understand the experimental research method
Overview of Consumer Behavior
Consumer behavior: the totality of decisions about
the consumptions of an offering by decision-making
units over time
How to get wise about consumers?
- Watch each decision-making stage (5 W’s)
o Who is the consumers?
o What is the consumer doing/looking
for/considering?
o When does the consumers make the purchasing decision, and what’s
going on before, during, and after?
o Where are consumers making decisions? (can vary by stage!)
o Why are they doing this?
- Asses the internal factors in the stage
o Human psychology: beliefs, attitudes, values, motives (drivers of
behavior), needs, wants, emotions (happiness, regret, anger, fear,
pride gratitude, shame, guilt), judgement and decision making,
information processing
- Consider the social factors
o Peers, other consumers, society: self-presentation, social
comparison, social norms, social motives, social influence tactics
- What aspects of the environment influence consumers? (external influences)
o Managerial, market, environment:
§ Market influences, economic factors: resource scarcity, economic cycles
§ Social and cultural factors: political climate, trends, events
§ Context eFects: time of the day, physical environment
§ Choice architecture (store lay-out and other factors)
,Lecture 2: Consumer Research / Studying the CSR Halo
Consumer Research
Surveys & focus groups are often inaccurate because:
- People don’t know what caused their behavior
- Memory is inaccurate
- People are bad at predicting future behavior
- People often don’t tell you what they know, because: social desirability, demand
characteristics, privacy/taboos
Scientific method: observe à question à hypothesis à method à result/conclusion
- Studying actual behavior overcomes most of this, that’s why we use experiments and
field studies
Correlational research strategy:
measures two (or more) variables for
each unit (consumer, firm, brand, etc.)
- E.g., relationship between ad
expenditures and sales
- Correlation is not the same as
causation
- Spurious correlation: two or more variables are
associated but not causally related to each other
o E.g., Is the number of crimes committed caused by
the number of ice creams eaten?
Basic elements of an experimental study:
- Manipulation: researcher manipulates one variable by
changing its value to create a set of two or more
treatment conditions
- Measurement: a second variable is measured for a group of participants resulting in
a set of scores in each treatment condition
- Comparison: the scores in one treatment condition are compared with the scores in
another treatment condition
- Control: all other variables are controlled to be sure that they do not influence the
two variable being examined
Terminology for experimental research:
- Independent variable (IV): the variable that is manipulated by the researcher by
creating a set of treatment conditions
o Condition: a situation or environment characterized by one specific value of the
manipulated variable (IV)
§ Levels: di\erent values of the IV, selected to create and define the treatment
conditions
- Dependent variable (DV): the variable that is observed for changes to assess the
e\ects or manipulating the IV
,- Extraneous variables: all other variables in the study other than the IVs and DVs
A/B tests: compares two versions of a (product/website/app/marketing message/etc.)
to see which one is the better performer
- E.g., Which color scheme will increase click rate?
o IV: color schema
o DV: click rate
o 2 conditions: blue & green
Within subjects design: single group of
participants
- Manipulates the IV by creating di\erent
conditions or treatments
- Measures the DV after each treatment
- Uses a single group of participants, who are exposed to each condition
- Advantage: no variation between subjects on other characteristics, so that your
statistical power is high
- Disadvantage: exposures can influence each other (this may be tackled by a
balanced or randomized presentation order)
Between subjects design: multiple groups of participants
- Manipulates the IV by creating di\erence conditions or treatments
- Measures the DV after each treatment
- Uses multiple groups of participants, each group is exposed to one condition
- Advantage: exposures cannot influence each other
- Disadvantage: participants may vary on other characteristics (this is tackled by
random assignment, which works better when groups are larger, so any
coincidental di\erences cancel out)
Minimizing variance within treatments
- Less variance = more precise and better interpretable results!
- Standardize procedures and treatment settings
- Random assignment and matching
o Limits confounding from individual di\erences
o Sample size: larger samples help reduce variance
- Minimize individual di\erences
o Create equivalent groups
o Reduce variance within groups
o You can use covariates to correct for remaining variance
- Random assignment allows for causal claims
- Random sampling creates representativeness
Analyzing results: statistical significance
- Develop a hypothesis about the e\ect of the treatment
- Conduct a test to examine if di\erence between groups (= the treatment e\ect) is
statistically significant
- What does a significant result mean?
, o The di\erence is large enough and consistent enough to rule out chance as a
plausible explanation (variance between groups is clearly larger than variance
within groups) à we can be reasonably sure (e.g., 95%) that the di\erence is
caused by the treatment
Applications and statistical analyses of between-subjects designs:
- Single-factor design with two groups
o Simples version of between-subjects experimental design
§ Compares the mean scores on DV of two groups of participants
§ Researcher manipulates one IV with only two levels
o Independent-measures t-test (or a single-factor ANOVA with two groups) tests if
the means are significantly di\erent
- Single-factor design with multiple groups
o Compares di\erent instances of a qualitative variable (e.g., color) or di\erent
levels of a quantitative variable (e.g., price)
o Overall e\ect is tested by a single-factor (“one-way”) ANOVA with multiple
groups
o Follow-up comparisons are needed to test di\erences
between specific pairs of groups/levels
Mediation: why does the effect occur? à mediation can be
addressed statistically by measuring an additional carriable
that captures the underlying psychological mechanism that
‘explains’ the effect
- NOTE: mediation is only correlational!
Moderation: what can change the effect? à academic
research uses multi (two-) factor designs to examine
moderation, moderators can be manipulated or measured
- From whom or in which situations does this e\ect occur?
Types of moderation
- No moderation: graph 1
o Only a main e\ect of IV (left panel)
o Main e\ects of both variables (right panel)
- Positive moderation: graph 2
o Strengthens the e\ect or makes it appear ‘necessary
condition’
- Negative moderation: graph 3
o Weakens the e\ect or makes it disappear ‘boundary
condition’
- Crossover moderation: graph 4
o Changes the direction of the e\ect: without main e\ects
(left panel), with a main e\ect of moderator (right panel)
Studying the CSR Halo
