REVISION OF SHORT QUESTIONS OF RESEARCH METHODOLOGY

REVISION OF SHORT QUESTIONS OF RESEARCH METHODOLOGY

• PARAMETRIC AND NON PARAMETRIC TEST
• IF THE INFORMATION ABOUT THE POPULATION IS COMPLETELY KNOWN BY MEANS OF ITS PARAMETER THEN STATISTICAL TEST IS CALLED PARAMETRIC TEST. PARAMETRIC OFTEN MEANS THAT A TEST ASSUMES THE TESTED VARIABLE ARE NORMALLY DISTRIBUTED. SUCH TEST OFTEN USE REGULAR NOMINAL VALUES AND AVERAGES
•  FOR EXAMPLE t-test,f-test,z-test and ANOVA
• IF THERE IS NO KNOWLEDGE ABOUT THE POPULATION OR PARAMETER BUT STILL IT IS REQUIRED TO TEST THE HYPOTHESIS OF THE POPULATION THEN IT IS CALLED NON PARAMETRIC TEST. FOR EXAMPLE MANN-WHITNEY,RANK SUM TEST,KRUSKAL-WALLIS TEST. INSTEAD OF NOMINAL VALUES . THEU LOOK AT THE DIFFERENCE IN MEDIAN INSTEAD OF AVERAGES

• DIFFERENCE BETWEEN PARAMETRIC AND NON PARAMETRIC TEST
• PARAMETRIC TEST

1. INFORMATION ABOUT POPULATION IS COMPLETELY KNOWN
2. SPECIFIC ASSUMPTIONS ARE MADE REGARDING THE POPULATION
3. NULL HYPOTHESES IS MADE ON PARAMETER OF THE POPULATION DISTRIBUTION
4. TEST-STATISTIC IS BASED ON THE DISTRIBUTION
5. PARAMETRIC TEST ARE APPLICABLE ONLY FOR VARIABLE
6. PARAMETRIC TEST IS POWERFUL IF IT EXIST

• NON PARAMETRIC TEST

1. NO INFORMATION ABOUT THE POPULATION
2. NO ASSUMPTION ARE MADE REGARDING THE POPULATION
3. THE NULL HYPOTHESIS IS FREE FROM PARAMETER
4. TEST STATISTICS IS ARBITRARY
5. IT IS APPLIED BOTH VARIABLE AND ATTRIBUTE
6. IT ISNOT POWERFUL LIKE PARAMETRIC TEST

• TYPE 1 AND II ERROR

1. WHEN A STATISTICAL HYPOTHESIS IS TESTED,THERE ARE FOUR POSSIBLE RESULTS:
2. THE HYPOTHESIS IS TRUE BUT OUR TEST REJECT IT
3. THE HYPOTHESIS IS FALSE BUT OUR TEST ACCEPT IT
4. THE HYPOTHESIS IS TRUE AND OUR TEST ACCEPT IT
5. THE HYPOTHESIS IS FALSE AND OUR TEST REJECT IT

• IF WE REJECT A HYPOTHESIS WHEN IT SHOULD BE ACCEPTED TYPE 1 ERROR
• IF WE SAY THAT HYPOTHESIS IS FALSE BUT OUR TEST ACCEPT IT. TYPE II ERROR
• DIFFERENCE BETWEEN TYPE 1 AND TYPE II ERROR
• TYPE 1 ERROR
•  THIS ERROR OCCURS WHEN WE REJECT THE NULL HYPOTHESIS WHEN WE SHOULD HAVE RETAINED IT. WE BELIEVE WE FOUND THAT A GENUINE EFFECT BUT REALITY THERE IS NOT ONE. THE PROBABILITY OF TYPE OF ERROR 1 OCCURRING IS REPRESENTED BY ALPHA AND SIGNIFICANCE LEVEL IS .05. IT MEANS THERE IS 5% PROBABILITY OF IDENTIFYING AN EFFECT WHEN ACTUALLY THERE IS NOT ONE.
• TYPE II ERROR
• WHEN WE FAIL TO REJECT THE NULL HYPOTHESIS. WE BELIEVE THAT THERE IS NOT GENUINE EFFECT WHEN ACTUALLY THERE IS ONE, IT IS REPRESENTED BY β AND THIS IS RELATED TO TEST OF POWER. COHEN PROPOSED THAT THE MAXIMUM ACCEPTED PROBABILITY OF TYPE TWO ERROR IS .2
• DESCRIBE CHI-SQUARE AS GOODNESS OF FIT

1. IS STATISTICAL MEASURE USED IN THE CONTEXT OF SAMPLING ANALYSIS FOR TESTING THE SIGNIFICANCE OF A POPULATION VARIANCE.
2. AS A NON PARAMETRIC TEST IT CAN BE USED AS A TEST OF GOODNESS OF FIT AND AS A TEST OF INDEPENDENCE OF ATTRIBUTE.
3. CHI SQUARE TEST IS APPLICABLE TO :
4. TEST FOR POPULATION VARIANCE
5. NON PARAMETRIC TEST

·         THE CHI SQUARE GOODNESS OF FIT IS APPROPRIATE WHEN THE FOLLOWING CONDITIONS ARE MET:

1. THE SAMPLING METHOD IS SIMPLE RANDOM SAMPLING
2. THE VARIABLE UNDER STUDY IS CATEGORICAL
3. THE EXPECTED VALUE OF THE NUMBER OF THE SAMPLE OBSERVATION IN EACH LEVEL OF THE VARIABLE IS  AT LEAST FIVE

• DESCRIBE CHI-SQUARE AS GOODNESS OF FIT

1. STATE THE HYPOTHESIS
2. FORMULATE AN ANALYSIS PLAN
3. ANALYSE SAMPLE PLAN
4. INTERPRET THE RESULT

• DIFFERENCE BETWEEN QUESTIONNAIRE AND SCHEDULE
• QUESTIONNAIRE

1. REFERS TO A TECHNIQUE OF DATA COLLECTION WHICH CONSISTS OF SERIES OF WRITTEN QUESTIONS ALONG WITH ALTERNATIVE ANSWERS.
2. FILLED BY THE RESPONDENT
3. RESPONSE RATE IS LOW
4. COVERAGE LARGE
5. LOW COST
6. RESPONDENT’S IDENTITY NOT KNOWN
7. SUCCESS RELIES ON THE QUALITY OF THE QUESTIONNAIRE
8. SUITABLE WHERE PEOPLE ARE LITERATE AND COOPERATIVE

• SCHEDULE

1. IS FORMALIZED SET OF QUESTIONS,STATEMENTS AND SPACES FOR ANSWERS PROVIDED TO THE ENUMERATOR WHO ASK QUESTIONS TO THE RESPONDENT AND NOTE DOWN THE ANSWER
2. ENUMERATOR
3. RESPONSE RATE IS HIGH
4. COVERALE SMALL
5. EXPENSIVE
6. COMPETENCY OF THE ENUMERATOR
7. USED IN BOTH LITERATE AND ILLITERATE

• KENDALL’S COEFFICENT OF CONCORDANCE ( W TEST)

1. IS NON PARAMETRIC STATISTICS. IT IS NORMALIZATION OF THE STATISTICS OF THE FRIEDMAN TEST. CAN BE USED FOR FOR ASSESSING AGREEMENTS AMONG RATERS. KENDEL W RANGE FROM ZERO ( NO AGREEMENT TO 1 COMPLETE AGREEMENT)
2. SUPPOSE FOR INSTANCE THAT A NUMBER OF PEOPLE HAVE BEEN ASKED TO RANK A LIST OF POLITICAL CONCERN FROM MOST IMPORTANT TO LEAST IMPORTANT. KENDELL’S TEST CAN BE CALCULATED FROM THESE DATA. ALL THE SURVEY RESPONDENT HAVE BEBN UNANIMOUS THEN W =1 AND W=0 AND THERE IS NO AGREEMENT.
3. KENDELL’S W MAKES NO ASSUMPTION REGARDING THE NATURE OF THE PROBABILITY DISTRIBUTION AND CAN HANDLE NUMBER OF DISTINCT OUTCOME
4. W IS LINEARLY RELATED TO THE MEAN VALUE OF THE SPEARMAN’S RANK CORRELATION COEFFICIENT. BETWEEN ALL PAIRS OF THE RANKING OVER WHICH IT IS CALCULATED,

• APPLICATION OF T DISTRIBUTION

1. IS A PROBABILITY DISTRIBUTION METHOD WHERE THE HYPOTHESIS OF THE MEAN OF SMALL SAMPLE IS TESTED. WHICH IS DRAWN FROM THE SYSTEMATIC POPULATION WHOSE STANDARD DEVIATION IS UNKNOWN.
2. IT IS STATISTICAL MEASURE USED TO COMPARE THE OBSERVED DATA WITH THE EXPECTED TO BE OBTAINED FROM A SPECIFIC HYPOTHESES

• APPLICATION OF T DISTRIBUTION:-

1. TEST OF HYPOTHESIS OF POPULATION MEAN
2. TEST OF HYPOTHESIS OF THE DIFFERENCE OF BETWEEN TWO MEANS
3. TEST OF HYPOTHESIZES OF THE DIFFERENCE BETWEEN TWO MEANS WITH DEPENDENT SAMPLE
4. TEST OF HYPOTHESSIS ABOUT THE COEFFICIENT OF CORRELATION

• APPLICATION OF Z TEST

1. THE Z TEST IS USED TO COMPARE MEANS OF TWO DISTRIBUTION WITH KNOWN VARIANCE
2. Z TEST IS A STATISTICAL PROCEDURE USED TO TEST AN ALTERNATIVE HYPOTHESIS AGAINST A NULL HYPOTHESIS
3. IT IS COMPARISON OF MEANS OF TWO INDEPENDENT GROUPS OF SAMPLES TAKEN FROM ONE POPULATION WITH KNOWN VARIANCE
4. Z= (MEAN OF THE SAMPLE-MEAN OF POPULATION)/STANDARD DEVIATION OF POPULATION AND n( no of observation)

• CONDITIONS

1. WHEN SAMPLES ARE DRAWN AT RANDOM
2. WHEN THE SAMPLES ARE TAKEN FROM POPULATION ARE INDEPENDENT
3. WHEN THE STANDARD DEVIATION IS KNOWN
4. WHEN NUMBER OF OBSERVATION IS >/=30