PSYC1001 Lecture Notes - Lecture 3: Sample Size Determination, Internal Validity, Stardew Valley
Research designs
Research designs
Anecdotes
• Interpreted stories about a single occurrence in the past and usually of NO scientific
value
• Small sample size
• Information passed down by non-scientists, share only certain details
• Up to interpretation
• No separation between theory and evidence
o Theory is only supported by evidence with is mentioned and vice versa
• Built-in bias
• Collecting anecdotes -> concentrating bias
• Can change with retelling
Case
studies
• More systematic
• All details recorded in scientific manner
• Arise from attempt to understand what is going on
• Small sample size
• Unbiased recording of information
• Never always linked to an explanation
• Scientific humility, objective view
Correlation
al studies
• Occur when at least 2 variables measured from each case/person, with a view to
calculating a relationship between the variables
• Measurements taken at same time, nothing is manipulated or controlled
• No intervention from experimenter
• Often difficult to know direction of causation or if there is evidence of causation
• Whether or not 2 studies are correlated
• Correlation =/= causation
• Advantages - allows researchers to determine strength and direction of
relationship, simple
• Disadv - no control over variables
• e.g. if variable x is correlated with variable y, then x could cause y, y could cause x
or var z could cause both
find more resources at oneclass.com
find more resources at oneclass.com
Screen clipping taken: 7/06/2017 2:42 PM
Screen clipping taken: 7/06/2017 2:42 PM
• Minimum - each correlation involves 2 scores from each case
• Correlation is description of how the dots are grouped
find more resources at oneclass.com
find more resources at oneclass.com
Document Summary
Interpreted stories about a single occurrence in the past and usually of no scientific value. Case studies: more systematic, all details recorded in scientific manner, arise from attempt to understand what is going on, unbiased recording of information, never always linked to an explanation. Small sample size: occur when at least 2 variables measured from each case/person, with a view to. Screen clipping taken: 7/06/2017 2:42 pm: minimum - each correlation involves 2 scores from each case, correlation is description of how the dots are grouped. Quasi experiment s: at least one key variable of interest cannot be randomly allocated, but others can, e. g. gender. True experiment s: all ivs controlled and able to be randomly allocated. Strong casual inference can be made as random allocation makes all variations between groups cancel pout. If difference is found in the dv you can conclude that differences in the iv caused that difference. Independent usually presumed cause in research: manipulated by experimenter.