15' Nonexperimental Research Designs

1
15. Non-experimental Research Designs

• Pelham Blanton
• Ch. 5

2
Case Studies(a more positive approach)

• Careful, in-depth analysis of an individual or
  group
• Usually focus on topics that are too
  extraordinary to studied in the lab
• Designed to generate, delineate, corroborate, or
  invalidate theories
• Usually carried out when little or no systematic
  research has been conducted

3
Abreu, P., João, P., Ribeiro, M., Forni, A.,
Pires, I., Sousa, G. (2005). Writing epilepsy
A neurophysiological, neuropsychological and
neuroimaging study. Epilepsy Behavior, 6, 463 -
466.

• Writing epilepsy is a rare reflex syndrome in
  which seizures are triggered by writing. We
  describe a 33-year-old, right-handed man, with a
  history of juvenile absence epilepsy in remission
  and a family history of epilepsy, in whom
  myoclonic jerks precipitated exclusively by
  writing started at the age of 30. Intensive
  video/EEG monitoring during neuropsychological
  tests revealed, at about 1 minute after starting
  to write, a dystonic posture, followed by
  myoclonic jerks involving the right hand that
  shortly after became generalized.

4

• Concomitantly, the ictal EEG documented
  generalized hypersynchronous polyspike-wave
  discharges, maximal over the right parietocentral
  area. SPECT revealed an ictal hyperperfusion and
  interictal hypoperfusion over right
  parietofrontal regions, and fMRI showed extensive
  and intense left frontal, supplementary motor
  area activation, induced by writing. This case
  study provides some evidence supporting the
  hypothesis that the mechanism underlying
  writing-triggered seizures may be a generalized
  seizure process, with a focal cortical trigger
  zone, presumed to be the left frontal lobe as
  suggested by clinical and fMRI data. A relevant
  role played by the right hemisphere (right
  parietofrontal region) is postulated in the
  full-blown expression of reflex epileptogenesis,
  as supported by EEG and SPECT findings.

5
Famous Case Studies

• Phineas Gage
• Brain damage left him with change in personality
  and deficits in reasoning.
• One of the first sources of knowledge that
  specific parts of brain associated with specific
  functions.

6

• Peter Tripp
• NYC disc jockey
• Attempted to remain wakeful for 200 hours
• Experienced horrible hallucinations
• But not Randy Gardner . . .?

7
K.C.

• Canadian!
• Are trait info and autobiographical info stored
  separately?
• K.C. had profound amnesia
• But could still know things about himself

8
Single Variable Research

• Census a body of data collected from every
  member of a population of interest
• Because it is so time and effort consuming to do
  a census researchers use population surveys
• Population Surveys survey of a representative
  portion of a population with the aim of
  generalizing to the entire population
• If use random sample you can typically make
  accurate estimates of a large population using a
  relatively small sample

9

• Cluster sampling The population is divided into
  clusters or groups, and some of these are then
  chosen at random. Within each cluster, units are
  then chosen by simple random sampling.
• e.g., conduct interviews with hotel managers in a
  major city about their training needs.
• Each hotel represents one cluster
• Randomly select a small number and contact the
  managers of these for interviews.
• one-stage cluster sampling When all units of the
  selected cluster are interviewed (e.g.,
  managers).
• two-stage cluster sampling If the subjects to be
  interviewed are selected randomly within the
  selected clusters. (e.g. instead of interviewing
  managers, you want to interview employees).

10
Sampling Error (Margin of Error)

• Sampling error the likely discrepancy btw the
  results one obtains in a particular survey and
  the results one would have likely found in the
  entire population
• e.g., Company X surveys customers and finds that
  50 of the respondents say its customer service
  is "very good." The confidence level is cited as
  95 3 percent. This information means that if
  the survey were conducted 100 times, the
  percentage who say service is "very good" will
  range between 47 and 53 most (95 ) of the time

11

• Sampling error decreases as the sample size
  increases (but only to a point)
• Two applications of population surveys are
• Epidemiological research
• Public opinion research

• http//www.raosoft.com/samplesize.html