Question? Leave a message!




Hypothesis Types

Hypothesis Types 3
Zainal A. Hasibuan  zhasibuacs.ui.ac.id)  RESEARCH METHODOLOGY COMPUTER SCIENCE FACULTY, UNIVERSITY OF INDONESIA Session Objec4ves  •  To understand the first step of conduc4ng  research: state general problem  •  To be able to iden4fy specific research  problems  •  To be able to construct hypothesis or research  ques4ons  •  To be able to determine research objec4ves Problem Iden4fica4on  •  Determine a well define general problem  –  Sources: research ar4cles, scien4fic discussion, experience,  etc.  –  Point of interest: phenomena, theories, previous research  results, etc.  •  Exercise your thought to express the problems  •  Write them as quickly as possible  •  Be ready, that some (may be all) of your general  problems are not researchable  –  Conduct literature search  –  Consult experts No Problem No Research  Agree  Or   Disagree Problem Formulation Problem Formulation Identification Research Problem Identification Scope of Research 1. Statistical Hypothesis 2. Hypothetical Statement Identification Research Questions Identification Research Objectives Hypothesis  •  Basically, there are to types of hypotheses:  – Hypothe4cal statement  – Sta4s4cal hypothesis  •  Hypothe4cal statement does not use  sta4s4cal analysis  – Apakah IT dapat meningkatkan kinerja  perusahaan  – Biasanya menggunakan pendekatan kualita4f  •  Sta4s4cal hypothesis uses sta4s4cal analysis.  – Menggunakan pendekatan kuan4ta4f.  Statistical Hypothesis   Twotail hypothesis: Reject H if calculated 0 value is either smaller than z ( t ) or greater α α than z ( t ) α α   Onetail hypothesis: Reject H if calculated 0 statistics value is greater than z ( t ) α α   Onetail hypothesis: Reject H if calculated 0 statistics value is smaller than z ( t ) α αStatistical Hypothesis Test: Steps 1.State the research hypothesis (“alternate hypothesis), H 1 2. State the null hypothesis, H 0 3. Choose an αlevel (alphalevel) Typically .05, sometimes .10 or .01 4. Look up value of test statistic corresponding to the αlevel (called the “critical value”)   Example: find the “critical” tvalue associated with α=.05 5.Use sta4s4cs to calculate a relevant test  sta4s4c.    – T‐value or Z‐value  – Soon we will learn addi4onal ones  6.Compare test sta4s4c to “cri4cal value”  – If test sta4s4c is greater, we reject H 0  – If it is smaller, we cannot reject H   0Hypothesis Test:  Errors  •  When we falsely reject H , it is called a Type I error  0 •  When we falsely fail to reject H , it is called a Type II  0 error  •  In general, we are most concerned about Type I  errors… we try to be conserva4ve  If I am false, If I am false, you are true you are true H a H 0 Taken in Combina4on, There are Four  Possibili4es  Researcher Accepts H Rejects H 0 0 Correct decision Type I error H is true 0 Probability = α Probability = 1 α Type II error Correct decision H is false 0 Probability = 1 β Probability = β Determine Research Objec4ves  •  Make sure you have a strong reason why such  research is worth of doing  –  Study the impact  –  Study the usage  –  Etc.  •  Research objec4ve should reflect the solu4on of the  problem  –  If you can not state the problem, you can not state the  objec4ve  •  Build your reasons based on facts (previous studies,  exis4ng problems, etc.)   •  (See slide: the aims of research) •  Write several general problems related to CS  and IT.  •  Build argument why such general problems  are worth to research.  •  Express your experience in sta4ng hypothesis  – Hypothe4cal statement  – Sta4s4cal hypothesis 
Website URL
Comment