Data Collection: Types, Sources, and Methods

Obtaining Data: Methods and Types

• Data comes from figures collected systematically for a predetermined purpose
• Data includes both quantitative and qualitative values of a variable for research

Types of Data

• Quantitative data measures values/counts and are expressed as numbers
• Qualitative data measures types and can be names, symbols, or number codes

Sources of Data

• Primary data originates from first-hand experiences, it has more reliability and authenticity, and is unpublished
• Secondary data refers to data that others have already collected

Data Collection Process

• Data collection involves a researcher gathering the information required to address a research problem
• Data collection factors: Which data to collect, how to collect, who will collect, and when data should be collected

Methods of Collecting Primary Data

• Primary data collection includes the observation method, interview method, questionnaire, and survey/schedule method

Observation Method

• Observation involves collecting information without asking questions.
• Advantages: No bias, current data
• Disadvantages: Time-consuming and subjective
• Divided into structured, unstructured, participant, non-participant, controlled and uncontrolled types

Interview Method

• Interviewing is a direct method of data collection via social interaction
• Advantages: In-depth information and flexibility
• Disadvantages: Can be expensive and time-consuming
• Types: Structured, unstructured, focused, clinical, group, qualitative, quantitative, individual, selection, and depth

Questionnaire

• Sending questionnaires to respondents contains a set of questions
• Advantages: Low cost, unbiased, and adequate response time
• Disadvantages: Low response rate and requires literate respondents

Survey/Schedule Method

• Involves a document with questions filled by enumerators
• Advantages: Clarifies questions and has a high response rate
• Disadvantages: Requires trained enumerators, time-consuming

Secondary Data Collection

• Sources of data include publications from government, technical/trade journals, books, magazines, newspapers, reports from industries/banks/stock exchanges, and reports from researchers/universities/economists/public records
• Factors to consider: Reliability, suitability, and adequacy of data

Sampling Methods

• Sampling identifies a population subset to accurately represent the entire group

Random Sampling

• Random sampling is picking respondents without design or order

Systematic Sampling

• Systematic sampling follows a set of rules to create regularity in the sampling process

Convenience Sampling

• Convenience sampling is the easiest method, but the least reliable, involves selecting the closest and easiest to reach participants

Clustered Sampling

• Clustered sampling involves using predefined groups and subgroups of populations

Stratified Sampling

• Stratified sampling divides a population into subgroups that share characteristics

Surveys

• Surveys are used to ask a lot of people well-constructed questions
• A survey contains a series of unbiased questions for the subject to answer

Steps for Designing a Survey

• Determine the survey goal
• Identify the sample population
• Choose an interviewing method
• Determine the ask and order of the questions and phrasing
• Conduct the interview and collect the information
• Analyze results by creating graphs and drawing conclusions

Planned Experiments

• An experiment refers to any process that generates sets of data

Experiment Terminology

• Response: The measurable outcome of interest
• Factors: Variables manipulated
• Treatment: Specified factor levels for an experimental run
• Replications: Systematic duplication of series of experimental runs
• Hypothesis: A supposition or proposed explanation
• Accuracy: Conforms to the correct value or a standard
• Precision: Measure of statistical variability to the true value
• Trueness: Closeness to the true value
• Control Group: The baseline measure
• Treatment Group: Item or subject being manipulated

Experiment Design

• A well designed and well conducted experiment is controlled, replicated and randomized

Randomized Design

• Experimental units are randomly assigned to either the control or treatment group

Block Randomization

• Involves placing subjects into groups of similar individuals

Scientific Method Importance

• The backbone of any experiment
• Is a process used to explore observations and answer questions, following scientific procedures

Statistical Methods for Experiment Design

• Includes analysis of variance (ANOVA), linear regression, and factorial design

Analysis of Variance

• ANOVA reveals statistically significant differences between samples/treatments

Linear Regression

• Linear regression models the relationship between a dependent variable and independent variable(s)
• Simple linear regression only has one variable
• While multiple linear regression has more than one

Factorial Design

• Factorial design evaluates how multiple factors affect a variable

Experiment Arithmetics

• Arithmetics involved in experimentation includes mean, error, percent error, deviation, and percent deviation

Mean

• The average of a set of numerical values

Error

• Error is the unknown difference between value retained and the true value

Percent Error

• An inaccuracy of a measurement, relative to the measurement size

Deviation

• Deviation measures the difference between the observed and another value, like the mean

Percent Deviation

• Percent deviation measures data points relative to the statistic average

Probability Definition

• Probability is a numerical description of how likely an event will occur
• . 0 on the scale is impossibility and 1 on the scale equals certainty

Calculating Probability

• Probability is the number of favorable outcomes divided by the total possible outcomes

Probability Terms

• P = Probability of an event to happen
• Q = probability to fail, p+q = 1

Experiment Definition

• Experiment is an outcome that cannot be predicted
• Sample space is the range of all possible expirement outcomes
• Every outcome in a sample space is an Element

Sets

• Complement: Subset of all elements not in set A
• Intersection: Event containing all elements common to sets A and B
• Disjoint: Sets A and B have no elements in common
• Union: Event that contains all elements in set A or set B or both

Venn Diagrams

• Venn diagrams graphically display probabilities of overlapping events

Combining Operations

• Operation-1 can be done in n1 ways, and if, for each of these, Operation-2 can be done in n2 ways
• Combine to get n1*n2 ways

Permutation

• Permutation is an arrangement of all parts of a set of objects
• Permutations are the number of ways to arrange a subset in a specific order

Distinct Permutations

• Distinct permutation of r objects taken from n
• Formula given as n!/(n-r)!

Cyclic Permutation

• The permutation of a objects in a circle is N=(n-1)!

Combination Distinct

• Combination of n distinct object is calculated as n!/r!(n-r)!
• Used to determine the ways to choose items without order

Probability

• Quantitative measure of how likely an event will occur

Probability Axiom

• For any event, the probability is more than zero

Probability Axiom 2

• 𝑃(𝑆) = 1 The sum of probabilities in a sample space is always 1

Probability Axiom 3

• If sets don't intersect, infinite collection is the sum of probabilities

Likely Outcomes

• The probability of event A happening is equal to one over number of outcomes

Independent Events

• Independent events is where probability of one event isn't affected by another

Exclusive Events

• Mutually exclusive events don't happen at the same time

Conditional Probability

• Is the probability based with the set sample space

Independence

• In independence, 2 evens remain the same regardless of other occurences
• P(A or B) = P(A) +P(B) - P(A and B)

Multiplication Rule

• This rules states of the product of probabilities of independent events occurring together

Random Variables

• Function on the sample space denotated by Capital letter

Discrete Variable

• Discrete means finite or countably infinite

Probability Function

• Probability of a random variable, is always between 0-1

Distributions

• Distributed Discrete Probability: distribution for random variables

Expectation

• Value is how much the values are spread out from the mean

Standard Deviation

• Standard is the sq root of the variation

Binomial

• Number of occurrences follows a fixed probability throughout trials

Distribution Contraints

• Trials are identical for times n
• Only two outcomes
• Consistant probability
• Events won't affect outcomes of others

Hyperthesis testing-Ronald A.Fisher

• Ronald was a English Statistician, Biologist, Genetics, and Eugenics knighted statistics

Jerzy Neyman

• Poland mathematician and innovator of neyman theory

Karl Pearson

• Karl was an english mathematician statistician

Egon S Pearson

• Karl's son and also an innovator

Hypotheses

• Test of hypothesis: includes null, alternate, test data etc

Hypotheses Type

• Nul Hypthesis accepted value
• Alternated hypothesis known and reserach hypothysis

Hypthesis Testing

• Outcome is based on results rejecting or failing to reject Test value is in critical regin

One Tail tests

• Test value is one side for either testing directions

Tails of tests

• In the null, is 𝑢=k.

Test statistic

• Reject or accept in value, with sample data
• Value is based on sample data and probability calculated using tests