BM CH3

• Variable


  something that varies, thus has at least 2 levels
• constant


  could potentially vary, but have 1 level in the study question
• Measured variable


  levels are observed and recorded
• manipulated variable


  variable that a researcher controls- assigning participants to the different levels of that variable

  
  

  Some variables can't be manipulated, only measured- i.e., age, IQ- impossible or unethical
• Construct (conceptual variable)


  a variable stated at an abstract level, usually defined as part of a formal statement of a psychological theory .
• Operationalising Variables


  turn conceptual definition into a measured variable or manipulated variable in order to conduct a research study

  
  

  E.G: Income= conceptual variable, operationalised= asking each person about their total income last year
• Claims


  argument someone is trying to make, researchers make claims on theories based on data, and journalist make claims when they report on studies they read in empirical journals

  
  

  
  

  Not All Claims Are Based on Research
• frequency claim


  rate or degree of a single variable ( 15% of American adults smoke- most ) - how frequency or common something is. Focus on one variable, variable is always measured not manipulated
• Association claims


  one level of a variable is likely to be associated with a particular level of another variables- correlate- one variable changes, the other variable changes.- related ( a late dinner is not linked to childhood obesity), at least 2 variables, links to correlational study, verbs: link, associate, correlate, predict
• positive association


  high levels of one variable go with high levels of another variable- y up, x up
• Negative association


  high levels of one variable, low levels of the other variable- as one goes up the others does down- y up, x down or x up, y down
• Scatterplot


  graphical representation of an association, where each to represents one participant in the study measured on two variables
• Causal claims


  one variable is responsible for changing the other, two + variables, two variables convay, you may see these claims based on zero association that reports a lack of cause (ie vaccines do not cause autism) , use language suggest that one variable causes the other- cause, enhance, affect, decrease, change. Advice is also a causal claim- if you do x, y will happen
• Move association to casual claim

  establish that the two variables are correlated, then must show that the causal variable came first and the outcome variable came later. Third, must establish that no other explanations exist for the relationship
• Interrogating the three claims using the four big validities
• Validity


  appropriateness of a conclusion or decision
• valid claim


  reasonable, accurate and justifiable claim
• Construct validity


  how well a conceptual variable is operationalised- how well a study measured or manipulated a variable
• external validity


  how well the results of a study generalise to or represent people or context beside those in the original study
• statistical validity


  study's statistical conclusion are precise, reasonable and replicable- how well do the numbers support the claim
• internal validity


  study's ability to rule out alternative explanations for a causal relationship between two variables.
• Construct validity of FC

  how well the researchers measure their variable of interest

  To ensure CV- ensure that variable has been measured reliably- measures create similar scores on repeated tested. Also ensure that different levels of the variable accurately corresponded to true differences in the variable
• External validity of FC

  Ask about generalisability- how the researcher choose the participant, do they represent the intended population and ask on external validity = how well the results of a study generalise to or represent people or context beside those in the original study
• Statistical validity of FC

  first start with point estimate- usually a percentage, next ask about precision of that estimate- precision captured by the confidence internal or margin of error of the estimate
• construct validity of AC

  Asses how well were the two variables measured- poorly measured- cant trust conclusions. However, good CV= confident in the association claim being reported
• External validity of AC

  Can generalise to other populations, contexts, time, places.
• statical validity of AC

  How strong the estimated association is( the strength of the relationship) and how precise that estimate is, and whether the study has been replicated in other studies- because multiple estimates are better than one
• Interrogating causal claims: three criteria for causation

  covariance

  
  

  temporal precedence

  
  

  internal validity
• Establish covariance


  the extent to which 2 variables are observed to go together- A changes as well as b changing-, established by the results of the study. High B high A, low b low a
• establish temporal precedence


  method was designed so that the causal variable comes first before the effect variable
• Establish Internal Validity


  study's method ensures that there are no plausible alternative explanation for changes in B- A is the only thing that changed - eliminated alternative explanations for the association
• Experiments Support Causal Claims

  to support CC- well conducted and designed experiment, where one variable is manipulated the other is measured. - construct validity
• A study's method can establish temporal precedence and internal validity

  manipulating and measure variables help make causal claims- helps ensure casual variable comes first (TV), manipulating a variable ensures a control over alternative explanations- ensure internal validity

  Methods of random assignment- increase internal validity

  
  

  Results can't achieve all four validities at once in an experiment- so they prioritize them- when making a casual statement means they may sacrifice external to ensure internal validity