Docstoc

3_HUL 204 Project work – Basic Concepts

Document Sample
3_HUL 204 Project work – Basic Concepts Powered By Docstoc
					Project work – Basic Concepts




                     Snehlata Jaswal


   HUL 204 LEADERSHIP, COMMUNICATION, AND DECISION MAKING IN ORGANIZATIONS
                  What is research?
Research is a systematic investigation of phenomena
and their relationships




Kerlinger (1986): Scientific research is a systematic,
controlled, empirical, and critical investigation of natural
phenomena guided by theory and hypotheses about the
presumed relations among such phenomena.
          HUL 204 LEADERSHIP, COMMUNICATION, AND DECISION MAKING IN ORGANIZATIONS
         Process of research
            Gap in                             Gap in
          knowledge                         knowledge:
                                            Framing the
         Reduced???                           problem




Hypothesis                                              Review of
testing and                                             literature
 statistical                                           and framing
 inference                                             hypotheses

                            Designing
                            the study
                               and
                            collecting
                              data
  HUL 204 LEADERSHIP, COMMUNICATION, AND DECISION MAKING IN ORGANIZATIONS
               Types of Research

Observational
     No control of variables is intended
Correlational
     Control possible only through selection of sample
     No manipulation of variables possible
Experimental
     Manipulation of independent variable(s) to study the
     effect on dependent variable
            Separate group design
            Single group design


       HUL 204 LEADERSHIP, COMMUNICATION, AND DECISION MAKING IN ORGANIZATIONS
                               Hypothesis
Every study – whether correlational or experimental – implies a
problem to be studied by the researcher.

The probable answer or solution to the problem is known as the
hypothesis.

A hypothesis has two essential qualities:
It is precisely stated so that it is testable.
It is not wild or improbable. It has a basis in previous research or
         theories.

            HUL 204 LEADERSHIP, COMMUNICATION, AND DECISION MAKING IN ORGANIZATIONS
              Types of hypotheses

Null hypothesis                        Alternate hypothesis

                                       One tailed (directional)

                                       Two tailed (not specifying a direction)




          HUL 204 LEADERSHIP, COMMUNICATION, AND DECISION MAKING IN ORGANIZATIONS
                     Levels of significance
Level of significance refers to the arbitrarily decided probability of
error accepted by a science.

The probability of error tolerated in any prediction varies from one
science to another. Since greater control and precise measurements
are possible in physical sciences such as physics and chemistry, they
test hypotheses at levels of significance as low as p < .001 or p <
.002.

Social sciences such as economics or sociology accept levels of
significance as high as p < .10. Psychologists generally test
hypotheses at either p < .05 or p < .01 (these may also be expressed
as respectively = .05 and = .01)
             HUL 204 LEADERSHIP, COMMUNICATION, AND DECISION MAKING IN ORGANIZATIONS
Errors in statistical decision making
Two types of error are always present in statistical decision-making
Type I: Rejecting the null hypothesis H0 when in fact it is true.
Type II: Accepting the null hypothesis H0 when in fact it is false.

                                       Decision                   Decision
                                       Reject H0                  Accept H0
              H0 true                Type I error                  Correct
                                        p=                         decision

              H0 false                  Correct                 Type II error
                                        decision                    p=



             HUL 204 LEADERSHIP, COMMUNICATION, AND DECISION MAKING IN ORGANIZATIONS
Errors in statistical decision making
 The probability of making a Type I error is directly indicated by .
 We are less likely to make an error of Type I if we have a smaller
 value of . The trouble is that as we decrease , we automatically
 increase – the chances of making Type II error – accepting the
 null hypothesis when it is actually false. Thus the researcher has to
 decide which of the errors is more acceptable. Since scientists
 abhor the Type I error more than the Type II error, they generally
 choose as small as .05 or .01.
 Type I error implies rejecting the null and accepting the alternate
 hypothesis to be true. Thus a wrong idea, hypothesis or theory may
 be accepted. The extremely cautious researchers prefer to make a
 Type II error. By accepting the null hypothesis when it is actually
 false, they leave the problem unsolved – as it existed before the
 statistical test. Thus they do not add any erroneous ideas / facts to
 the general body of knowledge that exists in the science.
          HUL 204 LEADERSHIP, COMMUNICATION, AND DECISION MAKING IN ORGANIZATIONS
Errors in statistical decision making
 Actually, the choice of should depend on the nature of the
 problem, and practical as well as theoretical considerations.

 For example, if a researcher is testing the effects of a new
 medicine, which may be harmful if taken in an overdose, a
 much smaller alpha is demanded. On the other hand, if a
 researcher tries to devise a test for identifying people at risk
 for committing suicide, such a test is useful, even if it makes
 very few correct predictions. A higher probability of error
 such as p < .10 may be quite acceptable in such cases.

 Thus, hypothesis testing is more a process of logical reasoning
 than statistical reasoning.
          HUL 204 LEADERSHIP, COMMUNICATION, AND DECISION MAKING IN ORGANIZATIONS
                         Measurement
Measurement implies assigning numerical values to
different qualities or quantities of a variable. So,
measurement may be qualitative or quantitative.

Levels of measurement (Stevens, 1951)
  Nominal scale
  Ordinal scale
  Interval scale
  Ratio scale
Most psychological variables are on an interval scale.
Non-parametric vs. parametric scales
          HUL 204 LEADERSHIP, COMMUNICATION, AND DECISION MAKING IN ORGANIZATIONS
           Scales of measurement
 Scale         Properties            Possible statistical operations


Nominal    Identity          Frequency
                             Percentage
                             Proportion
                             Mode
                             Coefficient of contingency
Ordinal    Identity          All above;
           Magnitude         Median
                             Rank order correlation
Interval   Identity          Almost all statistical operations possible;
           Magnitude         except coefficient of variation
           Equal intervals

 Ratio     Identity          All statistical operations possible
           Magnitude
           Equal intervals
           True zero
                                      Statistics
The term statistics has at least three different connotations. It refers to:
        – a body of knowledge; a subject of study
        – techniques and methods of treating data
        – data obtained from samples

Every researcher is interested in a population, but for practical purposes
we study only a sample of the population.

Measures that describe a population are parameters.
Measures that describe a sample are called statistics.

Parameters are fixed and true but hypothetical values.
Statistics vary from sample to sample and are thus prone to error. Yet
they are real values.
               HUL 204 LEADERSHIP, COMMUNICATION, AND DECISION MAKING IN ORGANIZATIONS
Techniques and methods of treating data

Descriptive statistics: Describe a sample. Necessary for
inferential statistics, but do not allow any conclusions/
inferences/ derivation of general principles and laws.

Inferential statistics: Help to draw conclusions from
the data. Predict parameters from statistics, and also
specify the probability of error associated with the
prediction(s).


          HUL 204 LEADERSHIP, COMMUNICATION, AND DECISION MAKING IN ORGANIZATIONS
                 Descriptive statistics
Measures of central tendency
      – Mean
      – Median
      – Mode

Measures of variability
      – Range
      – Average deviation
      – Quartile deviation
       – Standard deviation

Correlation



              HUL 204 LEADERSHIP, COMMUNICATION, AND DECISION MAKING IN ORGANIZATIONS
                    Inferential statistics
Correlation
– Product moment
– Rank order

t – ratio
– Independent groups
– Repeated measures

ANOVA
– One way
– Two way
– Multi-factorial ANOVA
– ANCOVA

FACTOR ANALYSES
REGRESSION ANALYSES
             HUL 204 LEADERSHIP, COMMUNICATION, AND DECISION MAKING IN ORGANIZATIONS
                    Thank you




HUL 204 LEADERSHIP, COMMUNICATION, AND DECISION MAKING IN ORGANIZATIONS

				
DOCUMENT INFO
Shared By:
Stats:
views:11
posted:6/12/2012
language:English
pages:17
Description: Leadership, communication and Decision is a must have course for every one. For a good Leadership you should have good communication skills and Decision making skills. These documents will teach you all that for free.