Week 1
4 elements of true experimental designs
1. Manipulation: researcher manipulates one variable by changing its value to create a set of 2+
treatment conditions (IV)
E.g. Condition 1 no sponsorship disclosure; condition 2: sponsorship disclosure
2. Measurement: second variable is measured for a group of participants to obtain a set of scores in
each treatment condition (DV)
E.g. Ad recognition (to what extent do children recognize the ad in videos?) measured in both conditions,
on a 7-point scale
3. Comparison: scores in one treatment condition can be compared across conditions (consistent
differences between treatment are evidence that manipulation caused changes; causality)
4. Control: all other variables are controlled to ensure they dont influence the two variables being
examined (IV effect on DV)
- Variable that changes randomly with no relation to IV is not a threat (random change not
systematic change)
- Hold variable constant
- Threat: Limit external validity
- Matching values across treatment conditions
- Counterbalancing!
- Randomization
- Random assignment (advantage: control multiple variables simultaneously)
Example 1: Milgram Experiment
Is the Milgram Experiment a true experiment? NOT A TRUE EXPERIMENT
Manipulation? NO: Every treatment was the same (all participants got the same treatment)
Measurement? YES: Severity of the shock (measurement of behavior = how far they would shock given
to students)
Comparison: NO: No manipulation/treatments/conditions there cannot be any comparison
Control? +/- : No manipulation makes it hard to say, but every student has the exact same experience
(e.g. voice of student)
Milgram’s experiment (obedience to authority) by Stanley Milgram
- Goal of study: How far would people go in obeying an instruction if it involved harming
another person?
, - Measured: extent to which participant would listen to authority
- Results: 65% continued to highest level
Example 2: The Marshmallow Test (impulsivity)
Toddles = subject to impulsivity (is it a characteristic they are born with? And can it be influenced by
their own logic?)
- Condition 1 Reliable: If they waited, they got better pens
- Condition 2 Unreliable : If they waited, the experimenter said they made a mistake and didnt get
better pens (had to use first option)
Results: Kids in condition 2 were more likely to eat marshmallow (mean: 3 min), in condition 1 kids
waited 4x longer (12min)
- Decisions on whether they should behave in long/short-term
Correlation vs. Causation
Goal of experimental research strategy: establish the existence of a cause-and-effect relationship between
2 variables
- Manipulate IV to create different groups which receive different treatments, the different scores
of DV are caused by treatment (IV) unless extraneous variables are involved.
Experimental Research Strategy
, Extraneous Variables aka. Control variables
- Extraneous variables: all variables beyond IV and DV
(goal of proper experimental design is to control these variables and
prevent them from becoming confounding variables)
Confounding Variables (problematic extraneous variables)
E.g. age/ time of day
When are they confounding variables?
1. Influences DV (if its totally unrelated, then its not a threat)
2. Varies systematically with IV (if it varies randomly, no threat)
(if it varies within conditions = threat!)
Methods to control extraneous variables
1. Randomization: unbiased procedure to
distribute extraneous values across conditions
(each research participant has an equal chance of
being in any of the treatments/ conditions)
Goal: minimize differences across groups (e.g.
distribution of gender/age is almost equal; no
individual differences between groups)
, 2. Holding a variable constant: e.g. if gender is considered a extraneous variable, only select only
males of 30 years all and divide them into the treatments
Internal validity low! (hard to generalize it to other groups)
3. Matching: systematic way of
dividing participants into treatment groups
(pair each similar participant based on age
& gender and one goes to treatment 1 and
the other in 2 = division is similar e.g. 70%
men, 30% women; not equal)
Question 1: Think about the difference
between correlation and causality (which was already in MCRS) and why we can test causal
relationships with an experiment but not with a survey. What's the difference? What element(s)
of a true experiment concern testing causal relationships?
Chapter 1 - Intro, Acquiring Knowledge & the Scientific Method
Methods of acquiring knowledge: ways in which a person can know things/ discover answers to
Qs
Scientific method:
Goal: obtain better-quality answers (or at least) a high level of confidence in the answers
Non-scientific approaches to acquiring knowledge
1. Method of tenancy: info is accepted as true bc it has always been believed/ superstitions
(as facts) support it (habit)
- The more we are exposed to these statements, the more likely we believe them to
be true
- Advertisers use the method of tenancy by repeating their slogans, hoping
consumers will accept them as true (& buy their products)
E.g. “ you can’t teach an old dog new tricks”; “ opposites attract”
- Problems: statements may not be true but there may be no ways of correcting erroneous
ideas. Even when faced with evidence that contradicts it, the belief persists (and is
difficult to change)
4 elements of true experimental designs
1. Manipulation: researcher manipulates one variable by changing its value to create a set of 2+
treatment conditions (IV)
E.g. Condition 1 no sponsorship disclosure; condition 2: sponsorship disclosure
2. Measurement: second variable is measured for a group of participants to obtain a set of scores in
each treatment condition (DV)
E.g. Ad recognition (to what extent do children recognize the ad in videos?) measured in both conditions,
on a 7-point scale
3. Comparison: scores in one treatment condition can be compared across conditions (consistent
differences between treatment are evidence that manipulation caused changes; causality)
4. Control: all other variables are controlled to ensure they dont influence the two variables being
examined (IV effect on DV)
- Variable that changes randomly with no relation to IV is not a threat (random change not
systematic change)
- Hold variable constant
- Threat: Limit external validity
- Matching values across treatment conditions
- Counterbalancing!
- Randomization
- Random assignment (advantage: control multiple variables simultaneously)
Example 1: Milgram Experiment
Is the Milgram Experiment a true experiment? NOT A TRUE EXPERIMENT
Manipulation? NO: Every treatment was the same (all participants got the same treatment)
Measurement? YES: Severity of the shock (measurement of behavior = how far they would shock given
to students)
Comparison: NO: No manipulation/treatments/conditions there cannot be any comparison
Control? +/- : No manipulation makes it hard to say, but every student has the exact same experience
(e.g. voice of student)
Milgram’s experiment (obedience to authority) by Stanley Milgram
- Goal of study: How far would people go in obeying an instruction if it involved harming
another person?
, - Measured: extent to which participant would listen to authority
- Results: 65% continued to highest level
Example 2: The Marshmallow Test (impulsivity)
Toddles = subject to impulsivity (is it a characteristic they are born with? And can it be influenced by
their own logic?)
- Condition 1 Reliable: If they waited, they got better pens
- Condition 2 Unreliable : If they waited, the experimenter said they made a mistake and didnt get
better pens (had to use first option)
Results: Kids in condition 2 were more likely to eat marshmallow (mean: 3 min), in condition 1 kids
waited 4x longer (12min)
- Decisions on whether they should behave in long/short-term
Correlation vs. Causation
Goal of experimental research strategy: establish the existence of a cause-and-effect relationship between
2 variables
- Manipulate IV to create different groups which receive different treatments, the different scores
of DV are caused by treatment (IV) unless extraneous variables are involved.
Experimental Research Strategy
, Extraneous Variables aka. Control variables
- Extraneous variables: all variables beyond IV and DV
(goal of proper experimental design is to control these variables and
prevent them from becoming confounding variables)
Confounding Variables (problematic extraneous variables)
E.g. age/ time of day
When are they confounding variables?
1. Influences DV (if its totally unrelated, then its not a threat)
2. Varies systematically with IV (if it varies randomly, no threat)
(if it varies within conditions = threat!)
Methods to control extraneous variables
1. Randomization: unbiased procedure to
distribute extraneous values across conditions
(each research participant has an equal chance of
being in any of the treatments/ conditions)
Goal: minimize differences across groups (e.g.
distribution of gender/age is almost equal; no
individual differences between groups)
, 2. Holding a variable constant: e.g. if gender is considered a extraneous variable, only select only
males of 30 years all and divide them into the treatments
Internal validity low! (hard to generalize it to other groups)
3. Matching: systematic way of
dividing participants into treatment groups
(pair each similar participant based on age
& gender and one goes to treatment 1 and
the other in 2 = division is similar e.g. 70%
men, 30% women; not equal)
Question 1: Think about the difference
between correlation and causality (which was already in MCRS) and why we can test causal
relationships with an experiment but not with a survey. What's the difference? What element(s)
of a true experiment concern testing causal relationships?
Chapter 1 - Intro, Acquiring Knowledge & the Scientific Method
Methods of acquiring knowledge: ways in which a person can know things/ discover answers to
Qs
Scientific method:
Goal: obtain better-quality answers (or at least) a high level of confidence in the answers
Non-scientific approaches to acquiring knowledge
1. Method of tenancy: info is accepted as true bc it has always been believed/ superstitions
(as facts) support it (habit)
- The more we are exposed to these statements, the more likely we believe them to
be true
- Advertisers use the method of tenancy by repeating their slogans, hoping
consumers will accept them as true (& buy their products)
E.g. “ you can’t teach an old dog new tricks”; “ opposites attract”
- Problems: statements may not be true but there may be no ways of correcting erroneous
ideas. Even when faced with evidence that contradicts it, the belief persists (and is
difficult to change)
