Research methods

what is Emperical research

Questions that can only answer by observation

Which are the 6 phases in decision making

1.Problem & need analysis

2.Find & design Options

3.Ex ante options evaluation

4.Choice

5.Implementation

6. Ex post choice evaluation

We use the word “Theory” for

For thinking

we use the word “Research Design” for

Planning

1.The process of start with Theory and thinking how we can test the theory , is called…

2. The process of start with Data and trying to arrive in conclusions with Data is called…

1.Deduction

2.inDuction

What type of question is this:

1.Why do some people start ?

1. Explanatory

What type of question is this:

1.How many people smoke?

2.Did tabacco companies do what they promised to do?

3. What are consequences of smoking?

1.Descriptive

2.Descriptive

3.Explanatory

An emperical research is clear if

Units and Variables are

What or who gives the…

units

If the Unit is known, ask “what” is …..

Variables

Normative questions are

What should be the case, often start with Should we ..

Conceptual questions

• What does it mean , often starts with “What is..”

• cannot be answer using observations

• based on agreement:talk together

What type of questions are the below:

1. Does the death penalty reduce crime?

Empirical-explanatory(causes or effects)

Questions with 2 variables called..

Casual

Two types of UNITS

1.Unit Observation

2.Unit of analysis

What is Ecological Fallacy ?

Drawing conclusions about lower level Units soley on the basis of aggregate data.

Example: Suicide rates in regions

What is Variable ?

Is a complete and mutually exclusive set of attributes or values.

5 levels of Measurement:

1. Dichotomy/Dummy

2. Nominal

3. Ordinal

4. interval

5. Ratio

What measurement is this ?YEs / NO

Dichotomy

If a variable has more than 2 attributes which are not ordered is called..

Nominal

If the values can be ordered but the distance between the values is unknow , the variable is called..

Ordinal

If the values can be ordered , the distance btw values is known , but there is no meaningful zero point the variable is called ..

interval

example: Degrees Celsius of a glass of water

If the values can be ordered the distance btw the values is known and there is a mmeaningful zero point the variables is called

ratio variable

Example: Amount of income in a household

Translate this to measurements.

1. Gender of a person

1. Dichotomy

A Codebook describes

The meaning of variables and values used in the data matrix

What is Independent variable?

A variable with values that are not problematic in analysis, but are taken as simply given.

What is dependent variable?

a variable assumed to depend on or be caused by another (independent variable)

What is Idiographic explanation ?

type of casual reasoning an approach to explanation in which we seek to exhaust the idiosyncratic cause of a particular condition or event, present specific cases fully

Wheel of science:

1. question

2. theory

3. research design

4. data collection

5. data analysis

6. answers/knowledge

Measuring theoretical constructs

Conceptualization—> operationalization—> measurement

What is Conceptualization ?

making clear what you mean with the theoretical construct. dimensions or aspects, and its attributes (values).

What is Operationalization?

the construction of the exact procedures to be used for data collection

What is content validity?

When all aspects of a construct are included in the operationalization

Data collection Methods?

• Survey

• diary methods

• Interviews

• Coding docments

• Physical measures

• Focus group recordinfs

Classifying data collection methods.

• Primary and secondary data

• From obtrusive to unobtrusive methods

• verbal and non - verbal methods

When the process is called obtrusive?

when the process of measurement affect the observations.

Verbal methods can be both unobtrusive and obtrusive.

Verbal methods can be both unobtrusive and obtrusive.

Verbal and Non-verbal methods?

• data based on or using natural language

• open interviews

• Focus group analysis

• ---

  Language can be misunderstood, misinterpreted.

  Verbal methods can be both unobtrusive and obtrusive.

  Example: coffee at the university (survey or listening to people near the coffee machine

How to measure a theoretical concept ?

• existing measurement instruments

  • articles and measurement

  • books with operationalizations(scales, indices)

• Thinking ,creativity

What is triangulation?

using at least two different operationalizations to measure the same theoretical concepts for the same units

what is Measurement ?

following a specified procedure (in the context of some data collection method) in order to ‘observations’, data .STORING DATA • data matri

STORING DATA

• data matrix:

  • rows

  • columns

What are the steps in survey research?

1. Are survey data avaliable(if yes , stop)

2. selecting the type of survey (web paper etc )

3. Designing and selecting survey questions

4. Designing the questionnaire

5. Questionnaire testing

6. interview training

7. survey administration - checking response

8. Data cleaning and storage

What is Meta data ?

Meta data is about data . it is a description of the data. Meta data explains how the data were collected, processed and published and answers questions regarding every facet of documented data

What can be the Drawbacks of using existing data & doing secondary data analysis?

• validity

• old data

Whats the 2 aspcets of data quality?

1. Reliability

2. Validity

What is reliability ?

The quality of measurement methods that suggest that the same data would have been collected each time in repeated observations of the same phenomenon.

What is external Validity ?

generalize to other settings

what is face validity?

The quality of an indicator that makes it seem to be a reasonable measure of some variable. That the frequency of atterndance at religious services is some indiacation of a persons religiosity seems to make sense without a lot of explanation.

What is criterion related validity?

the degree to which a measure relates to some external criterion.

What is Construct validity ?

degree to which a measure relates to other variables as expected within a system of theoretical relationships

What is content validity ?

The degree to which a measure covers the range of meanings included within a concept.

Two types of observation mistakes?

1. Random erros - reliability

2. systematic errors - validity

What is Reliability ?

Measurement reliability: the absence of random error (does everyone know what this means

and have an opinion about this?)

Measurement validity: no systematic error, (on average) measure what you intend to measure

Assessing the reliability – lack of reliability is a problem!

What is stability?

use the same operationalization several times, asking the same question 15 times to the same person, only small differences between observations

• a. test-retest reliability – asking the same question in the beginning and the end of a survey

• inter-rater reliability – asking several observers to rate/characterize/observe the same objec

What is consistency ?

• use and compare similar operationalizations

• items measure more or less the same thing, we expect the answers to be related

• strong relationship between variables (items)

Assessing the measurement validity

validity of an operationalization cannot be observed directly because teh true value is unknow . very difficult to measure

• measurement validity – does the measure reflect what is supposed to reflect?

• validity of research (strength of your conclusions)

Accurate observation

• precise (grams is better than kilos)

• reliable •

• valid

What is Univariate analysis?

the analysis of a single variable for purposes of description

E.g Frequency distributions ,averages , and measures of dispersion

What is frequency Distribution?

A description of the number of time the various attributes of variable are observed in a sample.

What do you know about the Center of Distribution?

• Mode-value that occurs most frequently, often used if a variable is measured in a nominal or ordinal level

• Median - middle value of your observation when they are ordered from the smallets to ge largest

• mean - sum of all the values divided by the number of observations , more sensitive to outliers

Summurazing data Frequently table (nimonal /ordinal)

• pie chart

• bar graph-number but nit percentage, better for high number of categories!

Quantitave data (interval /ratio)

• histogram (similar to bar graph)-

• boxplot

what is continous variables ?

a variable whose attributes form a steady progression, such as age or

income. Thus, the ages of a group of people might include 21,22,23,24 and so forth and could

even be broken down into fractions of years

What is Discrete variable ?

A variable whose attributes are separate from one another or discontinuous as in the case of gender or religious affiliation. In other words there is no progression from male to female in the case of gender.

What are the Measures of Variability?

• Variance- larger variance = larger variability, the more the values are spread out around the mean

• standard deviation-average distance of an observation frrom the mean , measure of dispersion, larger standard deviation= larger variability

Causality is not just correlation!

How to check time order?

Ex. Interrupted time series

design pre test -> pre test → treatment →post test → post test

What kind of measurement use for this ?

1. Country in which the headquarter of company is located

1. Nominal

what is Bivariate analysis?

the analysis of 2 variables simultaneously, for the purpose of determining the emperical relationship between them.

What is contingency table ?

• presenting the relationship among variables as percentage distributions.

• reveal the effects of the independent variable on the dependent variable.

• For nominal and ordinal variables.

What is multivariate analysis?

The analysis of the simultaneous relationships among several variables. Examine simultaneously the effect of age,gender and social class on religiosity is an example

What with testing bivariate hypothesis is checking?

• time order of cause and effect

• correlation /associationeffect of third variables

  • theorizing about the effect of third variables

  • formulating a trivariate hypothesis

  • testing the trivariate hypothesis

What is Cross-sectional study ?

study based on observations representing a single point in time

What is Longitudinal study ?

a study design involving data collected at a different points in time population is monitored over time

What is trend study ?

type of longitudinal study in which a given characteristic of some population is monitored over time

what is Cohort study?

study in which some specific subpopulation over time , although data may be collected from different members in each set of observations

What panel study?

Data are collected from the same set of people at several point in time

what is panel mortality ?

The failure of some panel subjects to continue participating in the study

what is pretesting?

The measurement of a dependent variable among subjects before they are exposed to a stimulus(erethisma) representing an independent variable

what is post-testing?

The remeasurement of a dependent variable among subjects after they have been exposed to a stimulus representing an independet variable

What is experimental group?

In experimentation, a group of subjects to whom an experimental stimulus is administered

What is a control group ?

a group of subjects to whom no experimental stimulus is administered and who resemble the experimental group in all other respects. thre is comparison of both groups at the end of eperiment, points to the effect of the experimental stimulus.

what is Double blind experiment?

An experimental design in which neither the subjects not the experimenter know which is the experimental groups and which is the control

What is randomization ?

a technique for assigning experimental subjects to experimental and control groups randomly

what is Matching ?

in connection with experiments, the procedure whereby pairs of subjects are

matched on the basis of their similarities on one or more variables, and one member of the

pair is assigned to the experimental group and the other to the control groups

What is external validity ?

possibility that conclusions drawn from experimental results may not be to the real world

What is statistical conclusion validity or correlation?

Correct conclusions drawn in the study itself

What is Internal validity ?

(is the time order between the measured variables correctly established? Is the relationship spurious)

Measurement validity and reliability

measuring the variables and units appropriately. Referring to the theoretical construct intended) à generalization & inference to theory/population/other cases

Measurement bias produces reverse causation

Measuring both variables at the same time, we don’t know whether dependent variable is dependent variable, and when we switch X with Y

What is Z-score ?

number of standard deviations removed from the mean. Useful to compare different distributions.

how is called the Recoding original scores into z-scores

standardization

What are the effects of using z-score for interval and ration variables?

1. standardizing variables using z scores change the center of the variable to 0

2. the standard deviation of a standardized variable is always 1

3. Standardization does not change of the distribution of the variable in any way , if it is oddly shaped it stays oddly shaped

what do you know about Empirical rule 68%-95%-99.7%

1. standard normal distribution is standardized,mean of 0 and standard deviation of 1. It is curve that results when any normal curve is converted to standardized scores

2. Normal distribution are unimodal and symmetrically distributed with a bell-shaped curve, however it can take on any values as its mean and standard deviation (for example mean might be 1000, and standard deviation 20)

Whats the 3 characteristics of an approximately normal distribution curve?

1. it has bell shape

2. mean and median are equal

3. 68%of the data falls within q standard deviation

4. population mena determines the location of the distribution

What is a parameter?

– a number that describes the data from a population

what is statistic?

a number that describes the data from a sample

Sample distribution VS Sampling distribution

for one single sample - – for lot of samples, of population.