Assignment  Quantitative Analysis
NOTE: There will be no pair/group work on this assignment. This is an individual assignment.
Introduction
We don't normally think of footwear as having user interfaces, and particularly, not audio interfaces. The data from this assignment is from an experiment with changing the audio interface of footwear and observing changes in user behaviour.
In this assignment you will analyse a dataset from a footwear user study. Participant gender, weight, height, and shoe size were collected. Participants experienced high frequency, low frequency and control audio feedback from walking while wearing the prototype shoes. For each of these states the researchers captured participant perceptions of their bodyweight, changes in their gait, and their mood, using the commonly used model of emotions consisting of the three dimensions: valence (positive/negative), arousal (calm/excited) and dominance (in control/overwhelmed).
Data Analysis
Your task is to analyse the quantitative data provided and write a report with your findings. You may use any tools you like for the analysis: Excel, SPSS, R, scripting languages, web based tools such as the online Adjusted Wald calculator, etc.
Requirements for Data Analysis
The data has already been cleaned and encoded (although you may find some issues or wish to encode things, or reorganise the data for analysis, in which case you should document what you have chosen to do). For your research you will need to:
1. Make decisions regarding missing or mismatched data, if relevant;
2. Calculate descriptive statistics on at least 5 data fields or records;
3. Quantitatively analyse and calculate confidence intervals to compare galvanic skin response for the three different frequencies (control, low and high). Can you conclude anything from your analysis?
4. Quantitatively analyse and calculate confidence intervals for the proportion of participants who had a positive emotional valence for the three frequencies. (You will need to determine the proportion of responses that were positive, i.e. greater than five on the ninepoint scale).
5. Quantitatively analyse the data to answer (at least) 3 other research questions that you want to answer from the data (for example, does perceived speed change with different audio frequencies?).
6. For maximum marks, at least one of the 3 other research questions that you choose should involve using data involving at least 3 variables.
Data Variables
The provided spreadsheet contains the following variables for each participant:
Variable Names/Name prefixes

Comments

Values

Gender

gender


Age

age


ShoeSizeUK

shoe size


Weight_Kg

body weight in kg


Height_cm

body height in cm


BodyVisualization_LOG

perceived body weight after each exposure, captured by the user altering an image of a
body until it matched their perceived body weight.

logarithmic scale

HeelPressure_Zscore

heel pressure

Z score

ToePressure_Zscore

toe pressure

Z score

FootAcceleration_Zscore

foot acceleration

Z score

GSR_Zscore

galvanic skin response

Z score

Valence

valence

9point scale (19)

Arousal

arousal

9point scale (19)

Dominance

dominance

9point scale (19)

Questionnaire_Speed

speed perception

7point scale (17)

Questionnaire_Weight

weight perception

7point scale (17)

Questionnaire_Strength

strength perception

7point scale (17)

Questionnaire_Straightness

body straightness perception

7point scale (17)

Questionnaire_Agency

perceived agency over
the heard sounds

7point scale (17)

Questionnaire_Vividness

vividness of body feelings

7point scale (17)

Questionnaire_Surprise

unexpected body feelings

7point scale (17)

Questionnaire_FeetLocalization

selfreported ability to localise feet.

7point scale (17)

Notes
1. Z scores are standard deviation scores after the data has been scaled (normalised) to have a mean of 0 and a standard deviation of 1. Recall that for normal distributions, ~95% of data will fall within two standard deviations, so most data points will be between 2 and 2. Z scores allow distributions of different types of data to be compared.
2. Participants experienced two trials for each case. That is, high frequency audio was experienced twice, as was low frequency audio and the control case.
Requirements for your Report
You will need to:
1. describe what you did to clean and encode the data (if anything)
2. state your decisions regarding missing data or mismatched data
3. state your decisions regarding the choice of statistical analysis techniques and parameters (eg. method of calculating confidence intervals, confidence level)
4. report descriptive statistics about the data
5. report on the analysis of your questions
6. include appropriate graphs in your report
7. summarise the results
8. clearly state any limitations relating to the results
9. clearly separate observations from conclusions
10. critically evaluate the experiment.
Report Presentation
To achieve maximum marks your work must fulfil these criteria:
 Have good use of layout and space.
 Be a professional report using appropriate language.
 Have sensible colours, fonts, and sizes.
 Be a maximum of FOUR pages long (excluding graphs and appendices). There will be a penalty for excessively long reports.
You may choose an academic paper format (eg. IEEE) or an industry report layout.
Attachment: Experimentdata.rar