Statistics Classification of Variables

- Branch of Mathematics that deals with the A. Qualitative Variables

collection, organization, presentation, analysis - Non-measurable characteristics that cannot
and interpretation of numerical data for the assume a numerical value but can be classified
purpose of making a meaningful decision. into two or more categories such as religion,
Branches of statistics gender and marital status.
1. Descriptive Statistics B. Quantitative Variables
- Concerned with the gathering, classification, - Exist in some degree along a continuum form,
summary calculations, graphs, charts and less to more and numbers can be assigned to a
tables. The summarizing values commonly used different individuals or object to indicate how
are the measures of central tendency, much of the variable they possess.
variability, and of skewness and kurtosis. 1. Discrete - variables that can take
- The purpose of a descriptive statistic is to tell specific or isolated values along a
something about the particular group of scaled. For example, counting the
observation. number of OFWs and the likes is
2. Inferential statistics considered a discrete variable.
- The logical process from sample analysis to a 2. Continuous - variables that are
generalization or conclusion about a obtained by measurement, usually with
population. It is also called statistical inference units and can be computed using a
or inductive statistic. formula such as the, the height and the
- Inferential statistics demands a higher order of weight of children, grades and
critical judgment and mathematical methods. It temperature.
aims to give information about large groups of
data without dealing with each and every Types of Data According to Source
element of these groups. It uses only a small Primary Data
portion of the total set of data in order to draw - are the data that are gathered directly from the
conclusions or judgments regarding the entire respondents of the study through observation,
set. interview, questionnaire, experiment or
measurement.
Population consist of all members of the group about Secondary Data
which you want to draw a conclusion. - are the data that have been previously
Sample is portion, or part, of the population of interest gathered, complied and are made available to
selected for analysis. the researcher for analysis. These include
Data are characteristics or information, usually books, journals, records, reports and other
numerical, that are collected through observation. In a publication.
more technical sense, data is a set of values of
qualitative or quantitative variables about one or more Methods of Collecting Data
persons or objects, while a datum is a single value of a 1. Interview
single variable - is a method of personal communication
Constant is a characteristic of objects, people, or events between the researcher and the respondents.
that does not vary. (For example, the temperature at This method provides consistent and precise
which water boils (100 degrees Celsius) is a constant.) information to the researcher because the
Variable a variable is a characteristic of objects, people, respondent may classify the information. It is
or events that can take of different values. It can vary in probably the most effective way of enlisting the
quantity (e.g. weight of people), or in quality (e.g. hair cooperation of the respondents.
color of people).
 Types of Interview Level of measurement
Structured Interview uses a research 1. Nominal Scale
instrument called interview schedule. - is the simplest and the most limited form of
It is made up of carefully prepared and measurement researchers can use.
logically ordered questions. - used to differentiate classes or categories for
Unstructured Interview is open and purely classification or identification purpose.
flexible method of interview. - The weakest form of measurement
The contents, sequence and wordings Example:
of the questions depends upon the researcher 1. gender (Male, Female)
who makes use of an interview guide, which is 2. religion (catholic, INC, Christians, etc.)
listing of topics that will be taken up during the 3. civil status (single, married, legally separated,...)
interview process. 2. Ordinal scale
2. Questionnaire - One in which data are not only classified but
- is a method where the subjects respond to the also ordered in some way high to low or least to
questions by writing or marking an answer most.
sheet. The questionnaire can be mailed or given - It is used in ranking.
to a large number of people at the same - It is somewhat stronger form of measurement
time. Examples:
3. Registration Method. 1. year level (1st year, second year, third year
- This method or gathering information is 2. faculty rank (Associate professor, professor,
governed by laws. (Example: birth certificates, instructor)
death certificates, and licenses, etc.) 3. military rank (general, colonel, private)
4. Observation Method. The difference between interval and ratio scales comes
- This method is used to data that are pertaining from their ability to dip below zero. Interval scales hold
to behaviors of an individual or a group of no true zero and can represent values below zero. For
individuals at a time of occurrence of a given example, you can measure temperature below 0
situation are best obtained by observation. One degrees Celsius, such as -10 degrees.
limitation of this method is that observation is Ratio variables, on the other hand, never fall below
made only one at a time or occurrence of the zero. Height and weight measure from 0 and above, but
appropriate events. never fall below it.
Participant Observation - Observations
done by actual participation, which allows the Sampling Designs
researcher to gain detailed and comprehensive Sampling
picture of the respondents. - refers to the process of selecting the subjects
Non-participant Observation - it is done who will participate in a research study.
where the participants are not aware that they are Types of sampling
being observed. It is advantageous for any kind of A. Probability sampling - is a process whose members
study since the participant will behave had an equal chance of being selected from the
naturally. population
5. Experiment Methods. This is used to determine the 1. Simple Random Sampling is one in which each and
cause and effect relationship of certain phenomena every member of the population has an equal and
under controlled conditions. This method usually independent chance of being chosen. Example is lottery
employed by scientific researcher. technique where each member of population unit is
assigned a number that is written on a slip of paper and
will be chosen one-by-one.
2. Systematic Sampling is a modified form of simple B. NON PROBABILITY SAMPLING- is a sampling
random sampling. It involves selecting every kth procedure where samples selected in a deliberate
element in one population until the desired number of manner with little or no attention to randomization.
samples is obtained. Also known as non-random sampling.
Example: Suppose in the list below, you are going to 1. Convenience Sampling this design is resorted to
choose every 5th on the list. when it is extremely difficult to select a random sample.
23 34 12 14 13 23 24 39 27 23 12 Example: A researcher may only include close friends
Therefore, the samples from every 5th from left to right and clients to be included in the sample population.
are 13 and 23 2. Purposive Sampling is a process of selecting based
3. Stratified Random Sampling is the process of from judgment to select a sample which the researcher
subdividing the population into subgroups or strata and believed, based on prior information, will provide the
drawing members at random from each subgroup or data they need.
stratum in the same proportion as they exist in the Example: A human resource director interviews the
population. qualified applicants in a supervisory position. (Note:
Example: Assuming that the total sample size is 3,455. Qualified applicants are selected by the HR Director
Field of Specialization Population which is based from his own judgment.)
Engineering 6,000 3. Quota Sampling is applied when an investigator
Business and Accountancy 500 survey collects information from an assigned number,
Total 6,500 or quota of individuals from one of several sample units
To determine sample size: fulfilling certain prescribed criteria or belonging to one
Step 1: population divide total population then multiply stratum. Their advantage is that they are cheaper to
to 100 administer.
Course Population Example: When the respondents are composed of men
Engineering (6,000/6,500) x 100 = 92% aged over 30 or 20 people who have bought cellular
Accountancy (500/6,500) x 100 = 8% phones in the last week. It is in the interviewer's
Step 2: Multiply percentage to the assumed sample size discretion which men or cellular phone buyers they
(3,455). select.
Course Population sample size 4. Snowball Sampling this design requires identification
Engineering 92% x 3,455 3,179 of a few persons whose qualifications meet the
Accountancy 8% x 3,455 276 purposes of the study. It is technique in which one or
Total 3,455 more members of population are located and use to
4. Cluster Random Sampling the selection of groups, or lead the researchers to other members of the
clusters of subjects rather than individuals. This population
sampling is design is used when the population is very Example: Imagine attempting to obtain the frame that
large and widely spread out over a wide geographical includes all homeless people in Metro Manila. To obtain
area. a sample of homeless individuals, for example, the
Example: If you wanted to know the opinion of Manila researcher will interview individuals on the street or at
regarding the effectiveness of the city government in homeless shelter.
their fight against the covid19, you can use cluster 5. Voluntary Sampling is a technique when sample are
sampling by subdividing the city into district then select composed of respondents who are self-select into the
at random the district to be used as sample. study/survey. Most of the time, samples have a strong
5. Multi stage Sampling is an extended version of interest in the topic of the study.
cluster sampling. It involves several strategies in Examples: Consider a news show asks their viewers to
drawing the representative sample from the population. participate in an on-line poll. The samples are viewers
The population units are grouped into hierarchy of who have chosen themselves and not the survey
elements, and sampling is done successively. administrator.
6. Judgment Sampling is a technique when the Classification - The process of putting together
researcher relies on his/her personal/sound judgment similar items from the mass of data collected
in choosing in participate in the study or the sample based on such characteristics as occupation,
selected is based on the opinion of an expert. sex, height, income, weight, nationality, etc.
Example: In a study where in a researcher such characteristics makes it possible to classify
wants to know what it takes to be a topnotcher data under particular headings.
in a bar examination, the only people who can Frequency Distribution Table - Refers to the
give the researcher firsthand advise are tabular arrangement of data by classes or
individuals who are bar topnotcher. categories together with their corresponding
class frequencies.
Presentation of data Steps in Constructing a Frequency Distribution
A. Textual Method - in this method, the writer can Step 1. Determine the range by getting the difference
emphasize the importance of some figures or can call between the highest and the lowest values in the set of
attention to the relevance of other figures. data.
Example: Range = HV - LV
The population in the Philippines increase from Step 1. Determine the number of class intervals or
106,651,394 in 2018 to 108,885,096 in 2019. the growth categories desired. The ideal number of class intervals is
rate is 2.09 percent. somewhere between 5 and 15.
B. Graphical Method - the data were presented using Sturge’s Formula: K = 1 + 3.3 log n
graphs or charts. Where:
Different kinds of Graphs/Charts K = is the number of class intervals
1. Line graph – it shows relationships between n = is the number of scores or observations
two sets of quantities. **Since k is number of class intervals, it is discrete.
2. Bar Graph – it consists of bars or rectangles Always use the next whole number regardless of the
of equal widths, either drawn vertically or decimal number.**
horizontally. Step 3. Determine the approximate size of the class
3. Pie chart – it represents relationships of the interval by dividing the range by the desired number of
different components of a single total as class intervals.
revealed in the sectors of a circle. Class size = Range/number of class intervals
4. Pictograph – it is a visual presentation of Step 4. Write the class intervals starting with the lowest
statistical quantities by means of drawing value for the lower limit of the first class. The upper
pictures or symbols related to the subject under limit is determined by adding the class size minus one to
study. the lower limit.
5. Map graph – one of the best ways to present upper limit of the first class = lowest value + (class size - 1)
geographical data. Step 5. Tally; Compute for the class mark by adding the
C. Tabular Presentation lower and upper limits of the class interval, then divide
It provides numerical facts in a more concise and the sum by 2.
systematic manner. It is constructed to facilitate the Step 6. To get the class boundaries, subtract 0.5 to the
analysis of relationships. Each class/subclass is assigned lower limit, then add 0.5 to the upper limit for each
to a particular row or column and figures for various class.
classifications are noted in appropriate cells. Step 7. To get the relative frequency, divide each
Tabulation - The process of condensing frequency to the total frequency then multiply by 100.
classified data and arranging them in a table. Step 8. To get < cumulative frequency, get the partial
Through this process, data can more readily be summation of the frequency for each class interval from
understood and comparisons may more easily the top.
be made.
Steps in Constructing a Frequency Distribution Step 6.
WITH EXAMPLE class boundaries
Class class
RAW DATA: interval boundaries
133 180 138 156 118 - 126 117.5 - 126.5
151 140 150 170 127 - 135 126.5 - 135.5
153 128 118 161 136 - 144 135.5 - 144.5
149 124 168 165 145 - 153 144.5 - 153.5
138 148 139 161 154 - 162 153.5 -162.5
142 147 167 130 163 - 171 162.5 - 171.5
143 137 147 146 172 - 180 171.5 - 180.5
150 149 129 142
158 152 130 175 Step 7.
148 142 159 120 Relative frequency
(2/40) * 100 = 5
Step 1. (4/40) * 100 = 10
Range = HV - LV relative
= 180 - 118 frequency frequency
R = 62 2 5
4 10
Step 2. 8 20
K = 1 + 3.3 log n; 14 35
= 1 + 3.3 log 40 6 15
K = 6.29 (use 7) 5 12.5
1 2.5
Step 3. 40
Step 8.
Class size = R/K
< cumulative frequency
= 62/7
118 - 126 = 2
= 8.86 (use 9)
127 – 135 = 2 + 4 = 6
136 – 144 = 2 + 4 + 8 = 14
Step 4 & 5.
upper limit of the first class = lowest value + (class size - 1) Continue…
Lowest value = 118 (lower limit of the first class) > cumulative frequency
class size = 9 118 – 126 = 40
= 118 + (9-1) = 126 127 – 135 = 40 - 2 = 38
The first class interval: 118 - 126 136 - 144 = 38 - 4 = 34
Repeat the procedure for the other class interval then Continue…
tally and compute class mark. Refer to the table below.
Class Frequency Distribution Table
interval frequency classmark Class interval frequency classmark class boundaries
relative frequency< CF > CF
118 - 126 2 122 118 - 126 2 122 117.5 - 126.5 5 2 40
127 - 135 4 131 127 - 135 4 131 126.5 - 135.5 10 6 38
136 - 144 8 140 136 - 144 8 140 135.5 - 144.5 20 14 34
145 - 153 14 149 145 - 153 14 149 144.5 - 153.5 35 28 26
154 - 162 6 158 154 - 162 6 158 153.5 -162.5 15 34 12
163 - 171 5 167 162.5 - 171.5 12.5 39 6
163 - 171 5 167
172 - 180 1 176 171.5 - 180.5 2.5 40 1
172 - 180 1 176
40
Class mark:
(118 + 126) / 2