Lecture 1 MLOS 500_0 060812

Category: Education

Presentation Description

No description available.


Presentation Transcript

MLOS 500a Research in Organizations:

MLOS 500a Research in Organizations Lecture 1 1

What is Research?:

What is Research? Question-oriented Planned and systematic Allows for serendipity Observation-based 2

What is Research?:

What is Research? Creative—looks for new questions Geared toward explanation Replicable—others may duplicate the process Self-critical—reports own flaws 3

What is Research?:

What is Research? • Public—open to questioning or examination by anyone • Cumulative and self-correcting—creates a shared history • Cyclical—ends with new questions to answer • It is a sound argument— claims are advanced on the basis of evidence • It is partial—there is always more 4

What is Research?:

What is Research? Research using the scientific method is different from other sources of “truth”, e.g., authority, experience, faith, reason, beliefs, etc. Research stands alongside, and sometimes in opposition, to everyday ways of knowing 5

Everyday ways of knowing:

Everyday ways of knowing Aren’t bad, but They tend to be unexamined They tend to be taken for granted We tend to remember anecdotal exceptions We think exceptions represent a bigger truth when they don’t People also tend to react to information based on how it is posed 6

A Typical Research Article:

A Typical Research Article 7

The Research Process:

The Research Process Assessment of relevant existing knowledge Formulation of concepts and propositions Statement of hypotheses Design the research to test the hypotheses Acquisition of meaningful empirical data Analysis and evaluation of data Provide explanation and state new problems raised by the research 8

Rules of the Research Process :

Rules of the Research Process Maintain “scientific objectivity” Work within your competence area Respect your subjects Give appropriate credit for others’ ideas, assistance, etc. Report conflicting evidence and all that you find 9

Rules of the Research Process :

Rules of the Research Process Describe the flaws in your research. Use primary sources whenever possible. Reporting of research must be done in a honest manner. When working in groups, give appropriate credit for work. Don’t take money for your work (conflict of interest) 10

Five basic ethical questions:

Five basic ethical questions Have the rules of scholarship been followed? Have the participants in the study been adequately informed and will they receive adequate debriefing? If deception is used in the study, is it appropriate? Is the topic being studied ethically appropriate? Has the researcher given thought to whose voice will be privileged in the research report? 11

Ethical obligations of the researcher:

Ethical obligations of the researcher The researcher must accept responsibility for the ethics of the research project Your participants must give informed consent to be in the project Deception must be justified, and must be immediately explained to the subject You must consider whether your subjects are likely to suffer psychological damage from participating in the project, and do your best to avoid harm 12

Defining the Problem:

Defining the Problem What is the problem? What are the variables? Where is problem located? Who is involved? When does it occur? 13


Variable any trait that can CHANGE VALUES from case to case 14

Variables Differ:

Variables Differ In how they’re measured Ordered/continuous variables—numerical values indicate how much of the concept is present—e.g., age, or communication apprehension Nominal/categorical variables—differentiated only on the basis of types or categories—e.g., kinds of strategies, sex, or race. 15

Variables Differ :

Variables Differ In how they behave Independent variable: Catalyst a variable that is expected to influence other variables a variable a researcher can manipulate Dependent variable: Result criterion or standard by which results are judged value is expected to be dependent on experimenter’s manipulation occurs as the result of the influence of the independent variable 16

Defining Variables:

Defining Variables Conceptual Definition What is it? Like a dictionary definition Operational Definition How do I know it when I see it? 17

Conceptual Definitions:

Conceptual Definitions Conceptual definitions describe what a concept means by relating it to similar concepts daily definitions—how most people define it scholarly definitions—specific, technical, used for a particular group of people 18

Conceptual Definitions:

Conceptual Definitions Problem is when daily and scholarly definitions get too far from each other Avoid circularity in definitions Best place to obtain conceptual definitions are handbooks, textbooks, specialized dictionaries (not Webster’s!) 19

Requirements of Conceptual Definitions:

Requirements of Conceptual Definitions Must include all situations or instances properly included in the term defined Must exclude situations or instances that are not properly included in the term defined 20

Requirements of Conceptual Definitions:

Requirements of Conceptual Definitions Must not use the term defined Must be more precise than the term defined Must exclude loaded language 21

Operational Definitions:

Operational Definitions The way a concept will be measured. Means of doing so Manipulated independent variables—researcher controls the variable—e.g., varying the amount of feedback an employee receives Measured/assigned variables—using attributes already in the situation—e.g., sex of respondent, personality characteristics, and controlling for their influence Direct classification variables—categorizing observable characteristics, e.g., observing people at customer returns desks and categorizing friendly/unfriendly behaviors 22

Standards for Operational definitions:

Standards for Operational definitions Observable and definite Logically consistent Inter-subjective—different researchers would classify them similarly Technically possible Repeatable 23

Other Important Terms:

Other Important Terms Data information collected in research projects that can be numerically represented 24

Other Important Terms:

Other Important Terms Primary Data = Data you get yourself Primary research = Research YOU get the data for Advantages: you get to decide variables you have total control Disadvantages: Much more time involved 25

Other Important Terms:

Other Important Terms Secondary Data = Data someone else has gathered or historical data Secondary research = Reanalyzing someone else’s data Advantages: You might have access to the best national data Saves time, effort and money Disadvantages: Have to live with other people’s definitions and questions 26

Time and Data:

Time and Data Cross sectional data analysis Gather all data at once (Example: Freshmen and seniors at the same time) Limitations on generalizability Most studies are cross sectional Longitudinal data analysis Gather data on a sample, wait till time passes, gather data on same sample. (Example: Freshmen now, back in four years for same group as seniors) If you’re studying what changes and why, this is better but often impractical 27

How data are measured:

How data are measured Nominal—classifications Sex, ethnicity, left-handed versus right handed Ordinal—classifications in an order Ranks of organizational members (admin assistant, manager, general manager, CFO, CEO, etc. Interval—assumes that the space between each measured item is equal Items on a scale Ratio—things that can assume a value of zero Profit, time, productivity 28

Data Example:

Data Example Horse races nominal : type of horse (chestnut, bay, palomino), color of silks ordinal : position of finish (win, place, show; ahead by 3 lengths) interval : how fast they ran last race, amount of weight they’re carrying ratio : what they’re worth 29

Reliability and Validity:

Reliability and Validity 30

Criteria for good measurement :

Criteria for good measurement No matter how well written a problem statement or literature review, or how insightful your analysis and conclusions, if you have lousy data you will have a lousy study . We use Precision , Reliability and Validity to check the likely accuracy of the data we collect. Precision and reliability are necessary but insufficient conditions for validity 31


Precision Precision has to do with the size of the measuring units Fractions vs. whole numbers Size of fractions 32


Reliability How accurate is the instrument that is being used? Am I getting consistent answers? Would a respondent answer the same from week to week? 33


Reliability Confirms that the outcome of a measuring process is REPRODUCIBLE. If you do it twice, you’ll get the same results . Measures the DEGREE OF CONSISTENCY that the instrument or procedure demonstrates: Whatever it measures, it does so consistently. 34

Examples of Reliability:

Examples of Reliability Test-retest reliability for consistency Inter-rater reliability—do people observing the same thing rate it similarly ? Inter-item reliability: Will 3-4 items “hang together” (consistent answer on similar questions by same respondent)? 35


Validity Confirms you’re measuring accurately. Refers to the quality of a data-gathering method that enables it to measure what it is designed to measure. 36

Types of Validity:

Types of Validity Criterion validity Does my measure correlate with other measures of the same construct/concept? Example: Two different measures of introversion would give similar results 37

Types of Validity:

Types of Validity Construct Validity Is related to operational definition : How are you measuring the phenomenon you are studying? Example : Anxiety and educational performance If you are using pulse rate to measure anxiety, you are only accessing part of the behavior and are really studying pulse rate and performance 38

Types of Validity:

Types of Validity Content/face validity How accurately do the questions asked tend to elicit the information sought? Is the instrument measuring what it is supposed to measure? 39

Types of Validity:

Types of Validity Internal validity Is the research design free from bias? External validity Can the conclusions drawn from the sample be generalized to other cases? 40

Validity is stronger than reliability:

Validity is stronger than reliability 41 You can have reliability without validity but you cannot have validity without reliability

Selecting Respondents:

Selecting Respondents Distinguish between Population—the total collection of all cases in which the researcher is interested Sample—a carefully chosen subset of a defined population. The ideal sample is both representative and economical 42

Getting Answers:

Getting Answers Data is collected to Answer a research question Test a hypothesis 43


Hypotheses A statement of relationship between two variables X will cause Y X is related to Y 44

Writing a Hypothesis:

Writing a Hypothesis Hypothesis is not a shot in the dark—you have to have a reason to make the hypothesis Must state relationships between variables Must be consistent with what is known in the literature Must be testable 45

Formula for a Hypothesis—a :

Formula for a Hypothesis—a People who are [high in/low in/characterized by/exposed to] VARIABLE X will have [higher/more/greater/less of] VARIABLE Y than others who are [low in/high in/not characterized by/not exposed to] VARIABLE X 46

Formula for a Hypothesis—b :

Formula for a Hypothesis—b There will be a [direct/positive/negative/curvilinear] relationship between VARIABLE X and VARIABLE Y 47

Hypothesis testing :

Hypothesis testing Researchers start out by hypothesizing that there’s a difference or a relationship between two groups or variables But they state their idea as if there is no difference (the “null hypothesis”) This is similar to the “innocent until proven guilty” premise in the courts 48

Characteristics of Hypotheses:

Characteristics of Hypotheses Hypothesis is basically a syllogism Major premise: If the theory is not true, then the hypothesis will not be supported by the data Minor premise: The hypothesis is not supported by data. Conclusion: Therefore, the theory is not true. 49

Science can be Weird:

Science can be Weird Even though it seems a little confusing, what we want to do is reject the syllogism shown before. That is, we want to find that the hypothesis is supported by the data, and therefore, the theory is true in this instance. WE CANNOT PROVE THEORIES TO BE TRUE. WE CAN ONLY TEST THEM REPEATEDLY. 50

Probability Estimates:

Probability Estimates When you read articles using statistics, they will often say something like The hypothesis was supported (p≤.05) or (p ≤.01) “P” stands for the probability of Type I error. P ≤.05 means that there are 5 in 100 chances that support for the hypothesis is in error P ≤.01 means that there is 1 in 100 chances that support for the hypothesis is in error 51

This should help you:

This should help you Think about a problem to research Critique your chosen article 52

authorStream Live Help