ATS2780 Lecture Notes - Lecture 6: Visual Thinking, Visual Communication, Semiotics
25 May 2018
School
Department
Course
Professor
Lecture 6 – Data, Distribution and Correlation
Visual methods
• Visual thinking
o Descriptive maps: presenting data
▪ Identify patterns and why these patterns occur
• Visual communication
o Demonstrative maps
Semiotics
• Ability to take apart signs and symbols seen on maps
• Iconic
o Looks like signified – means what it says
o E.g. coffee cup
• Indexical
o Need logical connection to thing
• Arbitrary
o Need to know what meaning is
• Sometimes symbols are used to produce bias maps
Surveys
• Types
o Spatial survey – e.g. grid, streams, etc
o Questionnaires to people
• Scope of survey
o Census
▪ Every member of the population – complete enumeration
▪ Ensure no sampling bias
▪ Can have census of streams
o Sampling
▪ Only a sample of the population - estimate
o Types of people
o Across streams
• Australian Bureau of Statistics
o Census every five years: population, housing, income, etc
▪ Every person must complete, referring to specific night
▪ Results released aggregated by local area l
o Monthly employment survey – based on sample population
▪ Sample of people surveyed
o Aggregate areas
▪ Blocks/streets/cities
Issues with Sampling
• Non representative sample
o Convenient sample – chosen by investigator
o Vast majority of subjects used for surveys are undergrad psych
students
▪ Young, healthy and reasonably well off
▪ Not a very good representative sample
find more resources at oneclass.com
find more resources at oneclass.com
Document Summary
Visual methods: visual thinking, descriptive maps: presenting data. Identify patterns and why these patterns occur: visual communication, demonstrative maps. Semiotics: ability to take apart signs and symbols seen on maps. Looks like signified means what it says: e. g. coffee cup. Indexical: need logical connection to thing, arbitrary, need to know what meaning is, sometimes symbols are used to produce bias maps. Intrinsic uncertainty: larger sample, more likely to get closer to correct answer, dependent on sample size, reported as -/+ value. Ethics in surveys and interviews: privacy of subjects is an ethical concern, survey subjects have an expectation of anonymity, watch for accidental reveals e. g. i survey the class"s rating of unit. I also ask each student"s course: but there"s only one student doing b. eco, don"t ask for home address, ask for postcode. Empiricism: gaining knowledge through the senses, there will always be some uncertainty, especially in sampling. Median: 50% of values will lie on either side of median.
