Transcript How is wine and cider preference correlated with self

Should You Ever Pay More Than $2 for a Bottle
of a Wine?
Eric Shed
Karen Thompson
Meredith Moran
Enrique Lopez
1) How much does an individual vary in their ratings of different wines?
2) Can this variation be used to determine wine quality?

Participants (N = 25) will be given four wine samples (A, B, C, D) of
predetermined varying quality (cost) and asked to rank each wine on the
following scale:

4 – Love it! (Great)

3 – Worth a 2nd glass (Good)

2 – Drinkable (Fair)

1 – Nasty! (Poor)


To address Q1, we will determine the mean and standard deviation for each
participant’s set of four scores. These SDs will then be averaged to
determine the average variability of the participants’ selectivity of wine.
To address Q2, we will use the T-score in an equation to determine the
quality of the wine.
Eric Shed, Karen Thompson, Meredith Moran, Enrique Lopez
Questions, Concerns, Observations



Q1 - Considering that we are only measuring variability of
preference across 4 wines, is this a large enough sample?
Q2 - We had the idea that we could somehow combine the Tscore with the individual’s overall rating (For example, if one
participant rated all wines the same, it would not indicate
difference in quality. All 4s are not created equal) to produce a
quality rating. Is this possible?!?!
Is it fair to use the relative cost of wine as a measure of its
quality?
Are people able to differentiate the
taste of different types of wine?
Daniel Stringer, Jason Murphy, Brian Edgar, Michelle Brown






Divide subjects into a control and experimental group.
The control group will taste from two glasses of the
same wine and will be asked if one tastes better or if
their tastes are of equal quality.
The experimental group will taste from two glasses of
different wines and will be asked the same.
We may re-run the experiment using multiple types of
wine
The researcher will then compare to see if the
experimental group is able to differentiate between the
wines more often than the control group.
Repeat process for cider.
Questions/Concerns


How can this process be changed to compare
all four varieties of wine?
What are some different ways to analyze the
data?
How is wine and cider preference
correlated with self-described level of
wine expertise?
Generalizability is dependent on the parameters of
the actual sample.
Nicky Ramos-Beban
Gloria Lim
Rachel Krefft
Janet Lundeen
Method



We will have people self-rate their wine expertise on a 4-point
scale

1-Novice (little to no experience drinking wine)

2-Low intermediate (can distinguish between grape varietals)

3-High intermediate (can distinguish between varietals and regions)

4-Expert (can distinguish between varietals, regions, and years)
We will have people taste each wine/cider and rate the taste on
a scale of 1-4 (1 being lowest, 4 being highest)
We will create a table comparing peoples’ responses to both the
expertise and taste questions and look for correlations between
the two variables.
Questions/Concerns

How to define “better” wines other than by taste
preferences (subjective)?
Hypothesis

Variance and standard deviation of experts’
preferences will be smaller than for other
groups
Research Question


How do Stanford EDUC 160 students’
preferences for red wine compare to
standardized Wine Spectator Ratings?
We will compare results using standardized
ratings from wine.com with EDUC 160
preferences.
Sabrina Peter
Nora Mallonee
Kevin Flynn
Jim Soland
Experiment



EDUC 160 students will complete a blind taste test of
wines
Each student will record their preferences (on a 5 point
scale) for four different characteristics (appearance,
fragrance, taste while in mouth, aftertaste)
Student preferences will be averaged and compared
with wine.com standard ratings, using the wine.com
ratings as the mean. We will use EDUC 160 data to
find the distribution of the data around the mean.
Observations


This experiment will allow us to determine the
degree to which EDUC 160 students’
preferences match with standard ratings of
wines.
Cheers!
Does the cost of a bottle of wine
correlate with people’s enjoyment of
that wine?
Ethan Hutt
Julie Cohen
Shayna Sullivan
Matt Kloser
Methods
1. Buy and record prices for each of the
several bottles of wine.
2. Conduct a blind taste test of each wine
with participants.
3. Participants rate each wine on a five-point
Likert scale (5=love it; 4=like it; 3=neither
like nor dislike; 2=dislike; 1=hate)
4. Find the mean likeability score. Compute
SD
5. Compare likeability score to price.
Questions, Interpretations


Can we record each person’s frequency of wine
consumption and compare it to how likely they
were to like or dislike expensive wines?
How do you compare two wines that may have
the same means but different SD’s? Exactly
what criteria can we use to rank the best
wines?
Do you get what you pay for?
The relationship between perceived quality
and price of sample red wines.
Elliott ‘pink lady’ Friedlander, Nick ‘spritzer’ Haisman, Seth
‘riesling’ Snyder, Luke ‘shiraz’ Terra, and Lambrina
‘Lambrusca’ Mileva
Experimental Method
We will evaluate the perceived quality of each wine using the following Likert scale:
1: Very Poor (suitable for minors)
2: Poor (gas station purchase)
3: Adequate (good 4th bottle)
4: Pleasant (weekday nights)
5: Spectacular (impressing in-laws)
Tasting method:

Blind tasting of each wine assigning values according to the Likert scale.

Tasters will cleanse palate (avec sorbet) between tastings.

Scores will be noted between samples.

Absolute silence will be maintained.
*Tasters will be selected at random from the participatory population, and will then be taken out of the population for further tasting
until every other member has had a turn. This will minimize the conflation of wine tastes and subsequent ratings.
Analysis:

Aggregate scores for samples will be divided by price per bottle to derive a pleasure per dollar measurement.
Limitations: These findings will be limited by the selection of only one type of wine, the sources of which might not be truly
representative. Price per bottle also does not include the whole cost of production including varying standards of environmental
Questions:


Are all bottles of equal volume?
Will only members of our group participate in
our sample, or the whole class?

How many ounces per sample?

Should each person taste wines twice?

Three times?

Spit or swallow?

Blindfolds? Berets? Candles?
Does the cost of a bottle of wine
correlate with people’s enjoyment of
that wine?
Ethan Hutt
Julie Cohen
Shayna Sullivan
Matt Kloser
Does the cost of a bottle of wine correlate with people’s
enjoyment of that wine?
Methods
1. Buy and record prices for each of the several bottles
of wine.
2. Conduct a blind taste test of each wine with
participants.
3. Participants rate each wine on a five-point Likert
scale (5=love it; 4=like it; 3=neither like nor dislike;
2=dislike; 1=hate)
4. Find the mean likeability score. Compute SD
5. Compare likeability score to price.
Ethan Hutt
Julie Cohen
Shayna Sullivan
Matt Kloser
Do you get what you pay for? The relationship between perceived quality and
price of sample red wines.
We will evaluate the perceived quality of each wine using the following Likert scale:
1: Very Poor (suitable for minors)
2: Poor (gas station purchase)
3: Adequate (good 4th bottle)
4: Pleasant (weekday nights)
5: Spectacular (impressing in-laws)
Elliott ‘pink lady’ Friedlander, Nick ‘spritzer’ Haisman, Seth ‘riesling’ Snyder, Luke
‘shiraz’ Terra, and Lambrina ‘Lambrusca’ Mileva
1) How much does an individual vary in their ratings of different wines?
2) Can this variation be used to determine wine quality?

Participants (N = 25) will be given four wine samples (A, B, C, D) of
predetermined varying quality (cost) and asked to rank each wine on the
following scale:

4 – Love it! (Great)

3 – Worth a 2nd glass (Good)

2 – Drinkable (Fair)

1 – Nasty! (Poor)


To address Q1, we will determine the mean and standard deviation for each
participant’s set of four scores. These SDs will then be averaged to
determine the average variability of the participants’ selectivity of wine.
To address Q2, we will use the T-score in an equation to determine the
quality of the wine.
Eric Shed, Karen Thompson, Meredith Moran, Enrique Lopez
people
Red wines
ciders
people
Red wines
ciders