Dr. Lee Revision Guide

Dr. Lee Revision Guide

1. Levels of Measurement

   • Nominal
     • Labels - Names
   • Ordinal
     • There is an order, but the difference between units is unknown
   • Interval
     • Ordinal + equivalent distances between units
   • Ratio
     • An infinite amount of numbers/quantity

2. Assumptions of inferential statistics

   • Normality
   • linearity
   • Equal Variance
   • multicollinearity
   • If assumptions are not met, then you cannot make assumptions on generalizability and others
   • statistics may provide valuable insight when they are meaningful, meaning when assumptions/conditions are met
   • Inferential statistics provide predictions about a population, based on data from a sample of that population - can predict characteristics of entire population by samples
   • Inferential statistical methods to research use sample statistics to make predictions about popular parameters

3. reliability & validity

4. Non-Probablity & Probability Sampling

   • sampling variablity and bias: sampling error, sampling bias, nonresponse error
   • sampling issues - does the researcher have the basis demographics information about the sample?
   • sampling errors : occurs when a sample does NOT well represent the population
     • Random Sampling error
       • increase sample size
     • Systematic Sampling Error
       • Reduce subjectivity
   • sampling terms
     • population
     • target population
     • accessible population
     • sample
     • census
     • parameter
     • statistic
     • sampling error
     • statistical relationship
     • Representative
   • Non-Probability
     • more common
     • based on the judgement of the researcher - as such, the selctive process is subjective
     • possibly subjective bias?
     • advantages:
       • Require less time
       • less expensive
       • more administrativley feasible
     • disadvantages
       • no way to analyse the probability of a unit of information, element, or individual being selected in investigation
     • types:
       • Purposive or judgemental
       • covenience
       • Quota
   • Probability
     • More generalisable
   • both approaches strive to be reresentative of the population and/or phenonmenom under study BUT employ entirely different methods to do so

5. Type 1 & Type 2 Error