applying science towards understanding behavior in organizations chapters 2 & 3
TRANSCRIPT
Applying Science Towards Understanding Behavior in Organizations
Chapters 2 & 3
Research Issues in Organizations
Approaches to collecting data Experimental Observational/correlational
Data collection issues Sampling
How should we select participants? What impact does it have on the results?
Experimental design Controlling potential confounds Assigning participants to experimental conditions
Measurement issues Describing and interpreting the results
Experiments: A Review
Experiments - Do changes in one variable (X) “cause” changes in another variable (Y)? Independent Variable (X)
condition or event that is manipulated by experimenter
Dependent Variable (Y) variable that is affected (hopefully) by manipulating
independent variable Extraneous Variable(s)
any variable other than independent variable that may influence dependent variable
Experiments: Pros and Cons
Advantage: Allows conclusions about direct effects of one
variable on another
Disadvantages: Experimental conditions are artificial
results may not “generalize” to the real world Some questions can’t be tested in an
experiment Require control that is not always available in the
“real” world
Experimental Design
Controlling potential confounds Goal of experiment is to “rule out” alternate
explanations of what affected dependent variable
Confounds are threats to internal validity Can be controlled through appropriate
experimental design and procedures
Internal Validity
History
Maturation
Testing
Instrumentation
Statistical Regression
Selection
Mortality
Selection-Maturation
Diffusion of Treatment
External ValiditySampleSetting (e.g., culture)Time (e.g., 60s vs. 90s)Replication (lack of)
Do the results of this experiment generalize (apply) to settings other than the experiment
Is there another reason (other than the independent variable) that could explain the results of the experiment.
Validity
How participants are selected for a study influences the extent to which the results can be applied to a larger group (external validity). A wide variety of techniques are available
Two Main types of sampling Probability
predetermined chance of any individual in the population being selected for the study
Nonprobability Typically nonrandom sampling
Sampling
Sampling Techniques
Probability Sampling1. Simple random sampling
2. Systematic sampling
3. Stratified random sampling
4. Cluster sampling
5. Multistage sampling
Nonprobability Sampling1. Convenience sampling
2. Quota sampling
3. Snowball sampling
Post with no Control Group
Training Posttest
Pre – Post with no Control Group
Pretest Training Posttest
Control Group with no Pretest
ExperimentalGroup
Training Posttest
Control Group Placebo Posttest
GroupDifferences
Pre – Post with Control Group
PretestExperimental
TrainingPosttest
Pretest Control Posttest
GroupDifferences
GroupDifferences
Measurement
Measurement – the process of assigning numbers to objects or events according to rules (Linn & Gronlund, 1995).
Psychological Measurement – concerned with evaluating individual differences in psychological traits. Trait – descriptive label applied to a group of
behaviors (e.g., friendly; intelligent)
Two basic types Descriptive
Describes the nature and properties of the data
Inferential Used in testing hypothesis
(e.g., differences between groups) (e.g., relationships between variables)
Data Analysis
Measures of Central Tendency
Measures of Variability
Distribution of the data
Descriptive Statistics
Measures of Central Tendency Mean
average score of all observations in distribution
Median midpoint of all scores in distribution
Mode most frequently occurring score in distribution
Descriptive Statistics
Measures of Variability Range
subtract the lowest from the highest score Standard Deviation
measure of the “spread” of the scores around the mean
Variance square of the standard deviation
Descriptive Statistics
Shapes of distribution curves Bell (normal distribution)
The bell curve has desirable statistical properties A number of inferential statistics “assume” data is
normally distributed
Skewed Curves Negative Skew - tail of the curve is to the left Positive Skew - tail of the curve is to the right
Distribution of the data
Properties of a normal distribution Measures of central tendency are the same
mean = median = mode
We know percentage of scores that fall within 1 standard deviation (68%) 2 standard deviations (95%) 3 standard deviations (99%)
Descriptive Statistics
Distribution in Normal Curve
The extent to which one variable can be understood on the basis of another Properties of correlation coefficient
direction (positive or negative) magnitude (strength of the relationship)
Cannot determine causality
Correlation
0
50
100
150
200
250
300
350
0 20 40 60 80 100 120
Exam Points
Fin
al G
rade
Poi
nts
r = .95
Scatter Plots (positive relationship)
0
50
100
150
200
250
300
350
0 20 40 60 80 100 120
Exam Points
Fin
al G
rade
Poi
nts r = .00
Scatter Plots (no relationship)
Job Satisfaction
Tur
nove
r In
tent
ions
r = -.95
Low HighLow
High
Scatter Plots (negative relationship)
Correlation: A Review