CS235 Lecture 5: Lec 5 - Oct 18
Midterm T/F and multiple choice
Keywords
● Survey formats
● Variables
● Measurement
● Operationalization
● 4 levels of measurement
National Public Radio
● Online survey conducted by dairy advocacy group
● 1000 American adults 18+ responded
● 7% of American adults think chocolate milk comes from brown cows
● Question worded: where does chocolate milk come from? However, we make chocolate
milk… question worded confusingly
● Survey lacks validity
Survey
● Method of collecting information from certain individuals/groups about their knowledge,
beliefs, attitudes, etc
● Typically makes use of discovery paradigm claims
● Typically aims to generate empirical, observable, and quantifiable data
● Can be used to describe a population
● Used to generate descriptive and causal claims
● Pros
○ Large populations
○ Large number of issues
○ Reliable (consistency in repeating study)
○ Time efficient
● Cons
○ Artificial/Unnatural
○ Not flexible/interactive
○ Close-ended questions limit expression
Administration
● Self-Administered
○ Individually or group administered (focus group)
○ Mail
○ Online
● Interviews
○ Face to face interviews, focus group
○ Intercept survey (mall intercept)
○ Telephone
○ Online
Questionnaire Format
● Open-ended - eg what do you like most about this course?
● Closed - eg are you a full time student? yes/no
● Filter and Contingency - are you a full time student? yes/no… if yes, how many courses
are you taking now?
Close-Ended Format
● Pros
○ Quick and easy
○ Better response rate
○ Easy to process data
○ Easy to compare answers
● Cons
○ Forced/coerced response
○ Potential errors in design → lack of measurement validity
○ Lack of depth
Structure
● Structured (standardized) interview
○ Each interview is presented in exactly the same questions in the same order →
high reliability
● Semi-structured interview
Document Summary
Online survey conducted by dairy advocacy group. 7% of american adults think chocolate milk comes from brown cows. However, we make chocolate milk question worded confusingly. Method of collecting information from certain individuals/groups about their knowledge, beliefs, attitudes, etc. Typically makes use of discovery paradigm claims. Typically aims to generate empirical, observable, and quantifiable data. Can be used to describe a population. Used to generate descriptive and causal claims. Closed - eg are you a full time student? yes/no. Potential errors in design lack of measurement validity. Each interview is presented in exactly the same questions in the same order high reliability. Follows a predetermined interview guide, with the opportunity to further explore responses. Questionnaire construction - types of questions, operationalization, and coding schemes. Operationalization: the process where constructs are turned into measurable variables. Hypothesis: playing violent video games increases aggressive behaviours among youth. Independent variables: to be controlled or manipulated, can cause corresponding changes in other variables.