Research Method.ppt

Dr. Shamanthakamani Narendran M.D. (Pead), Ph.D. (Yoga Science) R E S E A R C H M E T H O D S

D E F I N I T I O N S O F R E S E A R C H

Research is an active, diligent and systematic process of inquiry in order to discover, interpret or revise facts, events, behaviours, or theories, or to make practical applications with the help of such facts, laws or theories. The term "research" is also used to describe the collection of information about a particular subject.

Research means a systematic investigation, including research development, testing and evaluation, designed to develop or contribute to generalizable knowledge. Activities which meet this definition constitute research for purposes of this policy, whether or not they are conducted or supported under a program which is considered research for other purposes. For example, some demonstration and service programs may include research activities.

A careful hunting for facts or truth about a subject; inquiry; investigation. Scientific/technical research and development of new production processes or products, coupled with analysis and measures that provide information to potential users regarding the application of the new product or process; demonstration tests the feasibility of applying these products or processes via pilot plants and other pre-commercial applications.

The way in which the data are collected for the research project. All of the techniques, methods and procedures adopted in terminology work to carry out terminology research. DEFINITIONS OF RESEARCH METHODOLOGY

Statement of the Problem Purpose of the Study Objectives Review of the Literature Topic or problem area Theory area Methodology Methodology Subjects Design Data Collection COMPONENTS OF A RESEARCH PROPOSAL Data Analysis Bibliography Timetable

Brief description of topic In a paragraph or so, explain the topic you plan to explore for your literature review. Include a tentative title of your paper to bring greater focus to your topic. Rationale and significance Answer the question: "Why should I care?" Why is the topic you have chosen important to you? Research Questions To guide your literature search, what questions do you intend to answer. (One to three questions seem to provide maximum guidance, although, feel free to ask more if necessary.)

Tentative organization The organization of your literature review will in many ways depend upon your topic. Indeed, your outline may change dramatically by the time you turn in a final copy of the proposal. However, creating a general outline at this point can help focus your research efforts. A general outline will also help your instructor provide feedback.

Connection to Leadership Since your research paper should focus on systemic issues of leadership and not focus primarily on C&I (Curriculum and Instruction) (which is related primarily to classroom teaching/pedagogy), how will your research paper be substantively related to educational leadership?

H O W T O D O R E S E A R C H

HOW TO DO RESEARCH Identify an area for study which matches your competence. Discuss with experts about the avenines open for investigation. Identify a narrow area and a specific problem in it. Start a survey of the relevant literature, preferably with a review article. Literature survey can be manual or with a machine.

Manual survey is done by cross-referencing from article to article. Machine survey is through on-line data search. Both serve the same purpose, but the machine helps in saving time. Make notes of all papers and articles you refer. Literature survey helps in focussing on a specific research problem.

DESIGN THE RESEARCH STUDY What is the objective? What do you want to prove or disprove? What instrumentation is needed? If you do not have it, can you afford to buy it? If you have it, what is it you want to measure? Is the instrumentation reliable? Has it been calibrated? How often does it need to be calibrated and if so by what method?

How do you choose the control group and the experimental group? Do you need anybody else’s help and if so is it available? Are there any other groups in the world working on similar problems? If so how do you build contact?

Work out a time-frame for the research study. You may not be able to adhere to it, but at least it tells you what you wanted to do. Keep a diary of all measurements and observations on a daily basis. Do not trust your memory. It may play tries on you! Enter all data and observations in your PC/floppy/thumb file/CD and keep it with you. Numerical calculations can be done faster on a machine than manually.

M E A S U R I N G T H I N G S D A T A C O L L E C T I O N

MEASURING THINGS DATA COLLECTION Questionnaires Interviews Observation(s) Measurements

Questionnaires  2 times. Administered by trained interviewers. Self-administered. The former are considered (here) as STRUCTURED interviews.

Info elicited from a questionnaire gives/covers : [ What people __________. ] Attitudes “______ say they want” Behavior “_______ do” Beliefs “_______ think is true” Attributes “_______ are” There may be areas of overlap & other areas as well.

TYPES OF QUESTIONS Closed Open Organizational Dichotomous Scaled Check-list Funnel Filter Grid Ranking Freedom in way of responding The nature of the expected answer is indicated Responses are built up based on the first response

CLOSED QUESTIONS Answers are (to be) in a specified format. -- Dichotomous  two options YES/NO Main disadvantage – does not allow ‘shades of grey’ Eg. Has your ailment been treated (/managed) with yoga therapy? VERSUS Do you feel there is an improvement in your condition following yoga? Here wording of the questionnaire allows ‘ shades of grey ’

SCALED A scale is provided and the response is scaled. Do you take care to eat a healthful diet? OR Please show which (of the faces) indicates your mood?    Always Sometimes Never

CHECK LIST MULTIPLE CHOICE Respondents are asked to tick [  ] one out of many choices – sometimes one/more. Eg. Why did you choose yoga therapy? No other option – nothing else has worked! Your doctor suggested it Read about it and felt curious You know the importance of Mind-Body medicine today. [ Usually, no right or wrong ]

GRID / MATRIX More than one cell may have to be filled in, as relevant. Eg. Which of the following do you practice? MEDITATION KRIYAS PRANAYAMA ASANAS How many min(s)/day Tick if you practice

RANKING Subjects are asked to rank a list of things Eg. Arrange the following practices based on the amount of relaxation you feel, after them: Most / Midway / Least Cyclic meditation Bhramari (Py) “ OM” meditation Breathing exercises

OPEN QUESTIONS “Describe how you feel _________” Open questions are descriptive (a qualitative method of research) Organizational questions These questions Build on a response FUNNEL Direct the response FILTER

Which department do you work in? How many people work in that department? How long have you been working in that department? FUNNEL Do you practice yoga? [ If ‘YES,’ go to 2; If ‘NO,’ go to 3 ] How long have you been practicing? Do you practice any technique, thought to increase relaxation? FILTER Directs responses ?

What type of questionnaires are these : go back to 3 EXERCISE Which of the following symptoms do you have: What did you feel after practicing CM? Do you plan to return to PSK for a yoga therapy followup? Wheezing Shortness of breath How often/day Present (  ) or Not (  )

Best / Midway / Least How do you enjoy spending your free time? Walking Reading, watching TV Playing sports Sleeping (!) If yes go to (3), if no to (2) 1. Did you enjoy practicing CM? Have you found any other practice enjoyable? Have you felt a similar enjoyment with any other practice?

Hypothetical : [eg. Imagine that you are stranded on a desert island?] Difficult to imagine, may irritate. Hearsay [eg. What do you think X feels------?] Highbrow / intellectual. Long, rambling. Types of questions to avoid – differ based on the group / experimental setting

Check whether it is appropriate? A questionnaire is useful to cover large nos. Not for smaller nos. and in depth info. Characteristics of respondents. Define your domain [identify the research topic] When you design a questionnaire how do you ensure a good response rate ? Literacy ‘ First’ language (‘mother tongue) Failing eye sight

Presentation and layout? [eg. Is there enough space provided?] Instructions (as clear as possible) Other info [?] eg. To whom do I return the filled in questionnaire? eg. What is the purpose and applications of the study? 7. Distribution, followup [minimum length of time to return a questionnaire may be 2 wks]

Checking questions Results rendered in accurate: [ Ask same questions in a different way – two negatives ? ! ] Order / Position Where responses are scaled 20% of respondents use the middle category [ vary order / position ] How do you ensure that useful info is elicited

Differ for quantitative ques. (Even) open / qualitative can be analyzed: Which of the following do you enjoy? Logic and reasoning games Painting Playing a song based on written notes Learning formulae Right brain Left brain Check Analysis

Are ques. Understood? Are ques. Interpreted similarly? Reliability consistency on retest. Validity – how closely does it measure what it intended? <Reliability and validity to be detailed> Check accuracy Check reliability and validity

Access to large numbers. Data from closed ques. easy to analyze with statistics Advantages of questionnaires Low response rate 40% Insensitive to complex issues Questionnaire designing is complex and time consuming Disadvantages of questionnaires

An interview is a kind of conversation – a conversation with a purpose Face-to-face interviews allow you to observe nonverbal communication Interviews give more depth info (& hence more data !) than a quest. – which do you need? Interviews Structured Unstructured Telephone Types of Interviews Semi-structured Focus groups

Organize data Pictorially / Graphically Tabulate, before tabulation you may need to do Descriptive statistics ANALYZING DATA Bar / Column graphs Data curves Histograms Frequency polygons Pie diagrams Scatter grams Pict urograms Measures of central tendency Measures of dispersion

Usually tables contain RAW (= unprocessed) data with measures of central tendency & of dispersion. eg. Table : Memory scores in two groups On days 1 and 30 SD x D30 D1 Age Name D30 D1 Age Name CONTROL GROUP YOGA GROUP Sl. No.

Hypotheses may predict A relationship between 27 variables. The relationship to be either a difference / correlation and states the direction of the relationship. Inferential statistics go ---- further and test Hypotheses The direction Two (bi) One (uni)

Issues around research design Issues around TYPE of DATA Deciding on a test – inferential statistics Issues around research design eg. The experiment is to determine which instructions are easier to remember ( SHORT or LONG ) LEVELS The IV = instruction length The IV has 2 conditions or levels  Short or Long

VARIANTS The IV may also have >2 levels = Short / Medium / Long There may be >1 IV 2 There may be two treatments for mobility in children (Treatment A) (Treatment B) This gives a 2 x 2 design Adult Children Trt B Trt A group

Possibilities Different subject design (RCT, Xover, etc.) Same subject (Subject ) < Different subject design > Reduces the risk of ORDER effects Outcome influenced by several inter – individual variables. < Same subject design > Reduces difficulties related to inter – individual differences Difficulty – ‘ORDER’ effects ALLOCATION OF SUBJETS

< Matched subject design > attempts to reduce both difficulties Main problem : How perfect can matching be? Issues around the nature of data 2 points of importance Levels of measurement Distribution & variation of data

Nominal  eg. Males – 1 & females – 2 Ordinal  arranged in order but not absolute eg. Runners 1, 2, 3. Interval  arranged in order, equal intervals scale has no zero Ratio  order, equal intervals, zero Levels of measurement are 4

2 terms to be understood  Normal distribution  Homogeneity of variance Normal distribution refers to data in which the Mode Medium Mean Distribution & variation of data are the same

The bell shaped curve – normal distribution Z value = no. of standard deviations 95.5 99.7% 68.3 +3 +2 +1 -1 -2 -3

There is a statistical formula to determine if the distribution is homogenous / not (roughly homogeneity is where variability of scores in each condition are approx. same.) CI  the interval for P where P is the population possessing a characteristic for eg. CI 95 = 0.501, 0.599 means we are 95% confident that the interval 0.501 – 0.599 includes P

Back to inferential stats taking into account factors. IN SUMMARY These points form the basis – choose inferential stats RESEARCH DESIGN TYPE OF DATA No. of conditions No. of IV Subject allocation Levels Nominal Ordinal Distribution of data Homogenous Non-homogenous Interval Ratio Diff subj Same subj Matched subjects

Parametric are so called because various assumptions are made about the parameters of data Parametric tests are selected for INFERENTIAL STATISTICS 2 types Parametric Non parametric Interval or ratio data  essential condition Equal variance (both can be tested) Random – selected Non parametric  Ordinal data & non homogenous distribution

SOME QUESTIONS TO ASK PARAMETRIC – MORE POWERFUL Hypothesis Design Data Differences? Correlations? Same? Matched? Different? No. of levels? No. of variables? Data Distribution? Variance

FLOW CHART LOOKING FOR DIFFERENCES SAME, MATCHED, DIFFERENT? Same / matched Different How many conditions? 2 3 or more Variance ? homogeneity Variance ? homogeneity Equal Not Not Equal Parametric related T test Non-parametric Wilcoxon test Fried man test One way ANOVA Schutte test

DIFFERENT SUBJECTS HOW MANY GROUPS 2 3 or more Variance ? homogeneity Variance ? homogeneity Equal Not Not Equal Parametric Unrelated T test Unparametric Chi-Square Mann-Whitney test Kruskal-Wallis test One way Unrelated ANOVA

Flow chart looking for correlations HOW MANY GROUPS 2 3 or more Variance ? homogeneity Variance ? homogeneity Equal Not Not Equal Person test Spearman test Kendall’s coefft of concordance No definite test

A no. or TEST STATISTIC is obtained. The test stats. is looked up  standard tables  ‘P’ or probability. Probability – smaller values suggest significance ( eg. P <0.05, <0.001 ) Why? If P>0.05 – there is a 95% chance that results are due to chance P<0.001 ( the results being due to chance VERY unlikely ) AFTER RUNNING A STATS TEST

For a one tailed test there is a specific % probability that the difference might occur by chance If the hypothesis makes a bidirectional prediction then there is DOUBLE the prob. that the difference might be by chance 1 value in 1 direction + 1 value in 2 direction

G R A P H S D I S P L A Y D A T A

Graph  a GK word meaning “ to be drawn or written ” Typically, a graph is a pictorial representation of data. GRAPHS MAY BE USED TO DISPLAY DATA

The first step – (two) axes  one horizontal, other vertical Generally, the INDEPENDENT variable  X axis or abscissa and DEPENDENT variable  Y axis or ordinate Y axis ordinate X axis abscissa Abscissa Ordinate Categories of the variable (yoga, control groups) Frequency or relative frequency

Dependent – influenced Dependent – on the independent variables Hence a group clearly demonstrates relation between dependent variable & independent variable EDUCATION AND INCOME Income Education level (no. of years)

Bar graphs Data curves Scattergrams Graphing frequency distributions Histograms Frequency polygons Pie diagram Picturograms TYPES

DATA CURVES A, B, C  levels of education

GRAPHING FREQUENT DISTRIBUTION Histograms Frequency polygons Is a type of bar graph

PICTUROGRAM A B C A  Savings on expenses incurred due to polio B  Cost of immunization C  NET savings

Research Method.ppt

Related slideshows

More Related Content

What's hot

What's hot (20)

Viewers also liked

Viewers also liked (19)

Similar to Research Method.ppt

Similar to Research Method.ppt (12)

More from Shama

More from Shama (20)

Recently uploaded

Recently uploaded (20)

Research Method.ppt