Chapter 7 l Experimental Design The Laboratory Lab

Chapter # 7 l Experimental Design

The Laboratory (Lab) Experiment and The Field Experiment The Lab Experiment ~ Control. ~ Manipulation. l The Field Experiment l 2

The Lab Experiment l An experimental design set up in an artificially contrieved setting where controls and manipulations are introduced to establish cause and effect relationships among variables of interest to the researcher. 3

Field Experiment l An experiment done to detect cause and effect relationsip in the natural environment in which events normally occur. 4

Controlling The Contaminating Exogenous or “Nuisance” Variables Matching Groups. Ø Confounding characteristic. Ø Deliberately spreading them across groups. l Randomization: The process of controlling the nuisance variables by randomly assigning members among the various experimental and controls , so that the confounding variables are randomly distributed across all groups. l 5

Internal Validity l The confidence we place in the cause and effect relationship. 6

External Validity or Generalizability of Lab Experiments l The extent of generalizability of the result of a causal study to other field setting. 7

Trade-off Between Internal and External Validity l If we want high internal validity, we should be willing to settle for lower external validity and vise versa. 8

Factors Affecting Internal Validity l l l l History effects. Maturation effects. Testing effects. Instrumentation effects. Selection Bias effects. Statistical regression. Mortality. 9

History Effect l A threat to the internal validity of the experimental results, when factors unexpectedly occur while the experiment is in progress and contaminate the cause and effect relationship. 10

Maturation Effects l A function of the processes-both biological and psychological-operating whitin the respondents as a result of the passage of time. 11

Testing Effects l The distorting to subject the data collected to appropriate statistical tests, in order to substantiate or reject the hyphoteses developed for the research study. 12

Statistical Regression l The threat to internal validity that results when various groups in the study have been selected on the basis of their extreme (very high or very low) scores on the some important variables. 13

Selection Effects l The threat to internal validity that is function of improrer or unmatched selection of subjects for the experimental and control groups. 14

Mortality l The loss of research subjects during the course of the experiment, which confounds the cause and effect relationship. 15

Instrumentation Effects l The threat to internal validity in experimental designs caused by changes in the measuring instrument between the pretest and the posttest. 16

Types of Experimental Designs and Internal Validity l l l Quasi-Experimental Designs. True Experimental Designs. Solomon Four-Group Design and Threats to Internal Validity. Double-Blind Studies. Ex Post Facto Design. 17

Quasi-Experimental Designs Pretest and posttest experimental group design. l Posttests only with experimental and control groups. l 18

True Experimental Design Pretest and posttest experimental and control group designs. l Solomon Four-Group Design. l Double Blind-Studies. l Ex Post Facto Designs. l 19

Ethical Issues in Experimental Design Research Pressuring individuals. l Giving menial task and asking demaning questions. l Deceiving subjects. l Exposing participants to physical or mental stress. l 20

Ethical Issues in Experimental Design Research (cont. ) Not allowing them to withdraw from the research when they want to. l Using the research results to disadventage the environments. l Not explaning the procedures. l 21

Ethical Issues in Experimental Design Research (cont. ) Exposing respondents to hazardous and unsafe environments. l Not preserving the privacy and confidentiality. l Withholding benefits from control groups. l 22

Further Experimental Design Factor is used to denote an independent variable. l Level is used to denote various gradations of the factor. l Treatment refers to the various levels of the factors. l Blocking factor is a preexisting variable in a given situation that might have an effect on the dependent variable in addition to the treatment. l 23

The Completely Randomized Design Let us say that a bus transportation company manager want to known the effects of fare reduction, on the average daily increase in the number of passengers using the bus as a means of transportation. l The result of this study would provide the answer to the bus company manager’s question. l 24

Randomized Block Design l Through the above randomized block design, not only can be direct effect of each treatment, be assessed, but also the joint effect of price and the residential area route (the interaction effect). 25

Latin Square Design l The Latin Square Design is very useful when two nuisance blocking factors are to be controlled. 26

Factorial Design l The factorial design enables us to test the effects of two or more manipulations at the same time on the dependent variable. 27