Experimental Research Strategy

Chapter 7

Experimental Research Strategy

• the only research strategy that can determine cause-and-effect relationships between variables

Four basic elements of an experiment (have to satisfy all 4 to be real/true experiment)

1. manipulation of an independent variable

2. measurement of a dependent variable

3. comparison of groups

4. control of extraneous variables

Experiment conducted in Ancient Egypt

• wanted to find out why some prisoners did not die quickly from a Cobra bite

  • the prisoners that survived longer had eaten a citron (large lemon-like fruit)

    • does eating citron cause prisoners to survive longer after the bite?

Citron Experiment — Procedure

• any concealed citrons were confiscated when prisoners arrived

• prisoners were fed a standard diet for several days prior to their execution

• on execution day, prisoners were separated into 2 groups

  • one group were given a whole citron to eat before being bitten

  • the other group received nothing to eat

• all prisoners were then bitten

• the time it took each prisoner to due was carefully recorded

Citron Experiment — Results

• those prisoners who were not given a citron dies within minutes after being bitten

• those who were given a citron to eat survived for several hours before eventually dying

  • new rule that prisoners cannot eat citron before execution

Independent Variable

• the variable being manipulated by the researcher to test the effects on the dependent variable

Levels of the independent variable

• the number of conditions for a specific independent variable

  • ex; 2 levels — whole citron, no citron

Dependent Variable

• the variable being measured by the researcher

Controlling Extraneous Variables in an Experiment

1. hold them constant (environmental variables)

   • not always possible (ex; body weight, health…)

2. matching the levels of the variable across the treatment conditions (participant variables)

   • roughly the same in average health, body weight, age…

3. random assignment of participants to groups (participant variables)

   • each participant has an equal chance of being placed in any condition

   • the “great equalizer”

Categories of Extraneous Variables

1. Environmental (procedural) variables

   • ex; prisoners diet, size of snakes, location of snake bite…

2. Participant (individual difference) variables

   • ex; height, age, weight, health…

Experimental Group (treatment group)

• the group that is exposed to the independent variable in a study

Control Group (no treatment group)

• the group that is not exposed to the independent variable

Placebo Effect

• change in the dependent variable due to the participant believing (or expecting) a change will occur

Placebo Control Group

• participants are exposed to an inert, harmless substance that is like the independent variable but has no effect

  • ex; sugar pill, saline injection…

Manipulation Check

• an attempt to directly measure whether your independent variable had the intended effect on your participants (i.e., that your manipulation worked)

  • useful in pilot study to ensure your procedure is successful in creating stress before conducting the actual experiment so it can be modified if needed

  • can also be done in actual experiment but should be done at the end to avoid giving clues

Pilot Study

• a mini study that is conducted on a few participants prior to the actual collection of data (i.e., trial run)

Benefits to Manipulation Checks

• if manipulation check shows participants were not stressed then you’ve saved yourself the time and money doing the actual experiment

• if results are insignificant but manipulation check shows participants were stressed then you know the results are actually insignificant and its not an issue with the manipulation

• if results are insignificant and manipulation check shows participants were not stressed then the results of the study are not accurate because the manipulation did not necessarily because there actually isn’t a relationship

Laboratory Experiment

• an experiment conducted in a laboratory setting

  • allows studies to be done under highly controlled conditions

    • high in internal validity — single unambiguous result

    • low in external validity — laboratory setting is too controlled and different from naturally occurring events so the results cannot be generalized to the real world

Field Study/Experiment

• an experiment that is conducted in a natural (real life) setting (ex; on the street, in a school…)

  • high in external validity

  • Low in internal validity — likeliness for confounding variables in an uncontrolled environment

Mundane Realism

• the extent to which the materials and procedures used in an experiment are similar to something that the participants might encounter in everyday life

  • the extent to which an experimental situation resembles a real-life situation

Experimental Realism

• the extent to which the materials and procedures used in the study have an impact on and engages the participants

  • the extent to which an experimental situation makes the participants take the experiment seriously

Simulation Study

• the researcher attempts to create an environment like the one of interest to study people’s behaviour in a more realistic way

  • ex; Stanford Prison Experiment

Disadvantage of Simulation Study and Field Study

• you lose some control over the experimental study