Week 4

• Scientific thinking is based on...

  comparison, control , and manipulation.

  • a scientist compares conditions in the world. Without this comparison, we are left with isolated instances of observations, and the interpretation of these is highly ambiguous.
  • By comparing results obtained in different— but controlled— conditions, scientists rule out certain explanations and confirm others.
• Goal of experimental research

  • isolate a variable
  • the outcome of the experiment will eliminate a number of alternative theories that may have been advanced as explanations.
  • Scientists weed out the maximum number of incorrect explanations either by directly controlling the experimental situation or by observing the kinds of naturally occurring situations that allow them to test alternative explanations.
• Random assignment


  The subjects do not determine which experimental condition they will be in

  • randomly assigned to one of the experimental groups
  • scientists can rule out alternative explanations of data patterns that depend on the particular characteristics of the subjects.
  • ensures that the people in the conditions compared are roughly equal on all variables because, as the sample size increases, random assignment tends to balance out chance factors.
  • unbiased randomization device rather than the explicit choices of a human.
  • o systematic bias in how the subjects are assigned to the two groups
• Quasi experiment


  • comparison between groups
  • BUT lack manipulation (and some of the control involved in an experiment)
  • “naturally occurring” groups are compared – these groups are not directly created by a manipulation that is part of a study.
  • not “experiments” (i.e. “true experiments”)
  • If there is a “clear” difference between the groups, we may infer that we have a genuine difference between the groups, but we cannot rule out the possibility that something other than the defining feature of these groups is responsible for the group difference.
• Between-subjects design (true experiment)

  • participants andomly assigned to one of two or more groups (i.e., conditions), with the aim that the only thing that differs between conditions is the ‘level’ of the IV
  • These groups are then compared for the DV
  • by comparing the mean score of each group for the DV (if the DV is measurement data)/ by comparing the percentage of participants in a particular category (if the DV is a categorical variable).
  • If there is a “clear” difference between the groups, we infer that the manipulation of the IV had an effect on participants’ scores for the DV. 
• Field experiment


  where the IV is manipulated in a NON LABORATORY setting.
• True Experiment vs Correlational Study

  True experiment

  • manipulating IV hypothesized to be the cause and looks for an effect on the DV, holding all other variables constant by control and randomization.
  • removes third-variable problem in correlational studies (in the natural world, many different things are related)
  • isolates one particular variable (the hypothesized cause)
  • create special conditions that are unknown in the natural world.

  Correlational study

  • investigator simply observes whether the natural fluctuation in two variables displays a relationship.
• Representative sampling


  Each member of a population should have an equal chance of being selected during sampling

  • systematic differences in this chance produces a bias
• Sampling with vs without replacement

  • WITH: after a member of the pop is sampled, they are put back into the sampling pool to be sampled again (makes values independent)
  • WITHOUT: after a member is sampled, they cannot be sampled again
• Sampling error


  • determined by the quality of measurement made
  • sampling distrubution: value of our statistic will vary from sample to sample

  difference between the sample statistic and the true population parameter

  • differs from SD because it reflects variability between different sample estimates, whereas SD deviation measures variability within a single sample.
  • related to the sampling distribution as it defines how much sample statistics vary from the population parameter across repeated sampling.
• Standard Error of the Mean (SEM)

  SD of the sampling distribution

  
• SEM formula

  s= SD

  n = sample size

  
• Larger sample sizes yield...

  smaller SEM values
• Central limit theory

  
• PROBLEM SOLVING (two

  linear equations)

  
• Cramers rule