Practical Research 2

Types of Descriptive Research Design

1.1. Descriptive – Survey. This type of descriptive research design uses survey to gather needed data on a group of people without making any judgment.

1.2. Descriptive – Normative Survey. This type of descriptive research design is an extension of the descriptive – survey design. The inclusion of the word ‘normative’ means the gathered data (descriptive) must be critiqued to identify ways to improve it (normative).

1.3. Descriptive – Status Survey. This type of descriptive research design seeks to answer questions about real – life situations, with the assumptions that things may change.

1.4. Descriptive - Analysis Survey. This type of descriptive research design determines or describes the nature of an object by separating it into parts with the purpose of identifying or discovering the nature of things.

Descriptive – Classification Survey. This type of descriptive research design falls under the natural sciences. Its purpose is to classify animals and plants according to their unique traits or characteristics.

Example: A researcher collected various samples of mollusks from different

research stations and then identified and classified them accordingly.

1.6. Descriptive – Evaluative Survey. This type of descriptive research design seeks to carefully judge or assess the value of the topic being studied.

Two (2) Types of Descriptive – Evaluative Survey

1.6.1. Longitudinal Study. For example, a researcher conducted a study to evaluate the impact of Project Buhay on the self-reliance skills of junior high school students over a four-year period using the same group of subjects. The study started when the participants were in their 7th Grade (School Year 2012-2013) and finished when they were in 10th Grade (School Year (2015-2016).

1.6.2. Cross-sectional Study. For example, a researcher conducted a study to evaluate the effect of Project Buhay on the self-reliance skills of junior high school students. This was done simultaneously with students from all grade levels enrolled last School Year 2012-2013 as participants.

1.7. Descriptive – Comparative Study. This is a type of a descriptive research design which establishes a formal procedure to compare if a variable is better than the other (both are not researcher – manipulated) if significant difference exists

1.8. Correlational Study. This aims to describe and measure the degree of association between two or more variables or sets of scores.

2. Experimental Research Design. This design aims to accurately describe the facts and characteristics of a given population, situation, or phenomenon.

Types of Experimental Research Design

2.1. True Experimental. This type of experimental research design consists of three (3) characteristics: manipulation, control, and randomization.

2.1.1. Manipulation. An independent variable is being used to cause an effect to the dependent variable.

2.1.2. Control. There are at least two groups of respondents in a true experimental research design, the control group wherein the subjects do not receive treatment or intervention, and the experimental group where the subjects receive treatment or intervention.

2.1.3. Randomization. Every member of the population has an equal chance to be selected as a respondent. The selection may be through flipping of coin, draw lots, or computer-assisted random sequences.

2.1.4. Types of True Experimental Research Designs

2.1.4.1. Post-test only control group design

• This type of true experimental research design consists of two randomly assigned groups (experimental and control group).

• No pre-test/observation was administered before the implementation of treatment on the experimental group.

2.1.4.2. Pre-test post-test control group design

• This type of true experimental research design consists of two randomly assigned groups (experimental and control group).

• Pre-test and post-test were done on both groups before the intervention is carried out to the experimental group.

2.1.4.3. Solomon four group design

• This type of true experimental research design consists of four randomly assigned groups (2 experimental and 2 control groups).

• This is said to be the most credible research design since it minimizes the threat to external and internal validity.

• After the usual random assignment of subjects or respondents, experimental group 1 and control group 1 received the pre-test, followed by the intervention carried out to experimental group 1 and experimental group 2. After this, the post-test was administered to all four groups.

• Comparison was made among the four groups to assess the effect of the treatment/intervention/independent variable on the dependent variable.

2.2. Quasi Experimental Research Design. Just like True Experimental Research Design, this type of research design also involves the manipulation of independent variable to make an effect to a dependent variable. However, this design lacks at least one of the three characteristics that a true experimental research design has. This may either be randomization or control.

Types of Quasi Experimental Research Designs

2.2.1.1. Nonrandomized control group design or nonequivalent control group design

• This is like the pre-test – post-test control group design but participants were not selected randomly.

• Illustration below describes the design, wherein each group was given the pre-test (O), and then the intervention (X) was carried out to the experimental group, and then finally the post-test (O) was given to both groups.

2.2.1.2. Time series design

• This type of quasi experimental research design is ideal for studies that require to measure the effects of the treatment for a long period of time.

2.3. Pre-experimental Research Design. This type of experimental research design is considered as the weakest among the three types of experimental research because the researcher has very little control over the experiment and certainly has no control over threats to internal validity.

Types of Pre-experimental Research Designs

2.3.1. One shot case study. This type of pre-experimental research design has no randomization, no control group, and no pre-test. After the experimental group has been exposed to a treatment, a post test was administered.

2.3.2. One group pre-test - post-test design. This type of pre-experimental research design has no randomization and no control group. A pre-test observation was given before the implementation of the treatment, followed by a post-test observation to assess the effectiveness of treatment on the subjects.

2.3.3. Static group comparison study. This type of pre-experimental research design consists of two groups, one experimental and one control. A post-test was administered after the treatment to measure the significant difference if there is any. No pre-test was done to both groups.

Lesson 2. Sampling Procedure and Sampling

RESEARCH-tionary! Read and study the following terms for a more guided discussion of this topic.

A. Population refers to the entire group that the researcher wants to study.

B. Population size refers to the number of subjects in a population and is usually represented by “N”.

C. Sample refers to the specific group that either serves as the representative of the entire population or have met a certain set of

qualifications. This is where the researcher will collect the data from.

D. Sample size refers to the number of subjects included in a study and is usually represented by “n”.

E. Sampling technique refers to the process by which the samples have been selected.

F. Subject refers to the individual participating in a research study. It is

9 | P a g e

also known as case, participant, or respondent.

2.4. Population vs. Sample

In instances where the researcher does not have enough time or resources to make the entire target population as subjects/respondents, he/she may select a few from the whole population to participate in his/her study. This group of subjects is called the sample.

But what are the various ways of selecting these subjects?

Sampling Technique

2.4.1. Probability Sampling. This is a type of sampling technique where all the members of the population have equal chances to be selected as subjects. This will be done through the process of randomization.

Types of Probability Sampling

2.4.1.1. Simple Random Sampling

• This allows the researcher to select his/her subjects through drawing lots, using a table of random numbers or other random number generators.

• Do this until you achieve the required number of subjects.

• This is to ensure that the selection will be based entirely by chance.

10 | P a g e

2.4.1.2. Systematic Random Sampling

• This uses intervals to select the subjects.

• Find the total number of the target population and divide it to the required sample size (if there are 500 students and you need 100 subjects, divide 500 to 100 and the answer is 5).

• Get a list of all Grade 8 students arranged in any order (Note: Avoid list that is influenced by any bias towards any of the subjects.)

• Next, choose any number from 1 to 10 (for example 4). From number 4 onwards, every 5th person on the list will be selected (4,9,14,19…).

• Do this until you achieve the required number of subjects (in this case the sample size is 100).

2.4.1.3. Stratified Random Sampling

• This divides the population into groups (stratum/strata) based on their profile (it can either be according to age, socio-economic status, gender, or anything depending on the study).

• The purpose of this sampling technique is to obtain a sample population that will best represent the entire population being studied.

• To identify the subjects, start by getting the total number of the target population.

• Determine the sample size.

• Get the total number of individuals in each stratum.

• Divide each number of individuals in each stratum by the total population then multiply by the desired sample size.

• In the sample where the total population is 500, 300 males and 200 females, with 50 sample size, divide 300 males by the total population of 500 then multiply by the desired sample size of 50 to get 30. Same will be done to females. 200 females divided by total population of 500 multiplied by the sample size of 50 to get 20.

• So, there will be 30 male and 20 female participants with a total sample size of 50.

• Simple random sampling or systematic random sampling may then be used to identify the subjects in each stratum.

2.4.1.4. Cluster Sampling

• Just like stratified random sampling, cluster sampling is also divided into subgroups. However, they are different in the sense that:

a. cluster sampling is used when the target respondents are spread across a geographical location (province, barangays, schools, etc.). Instead of strata, the groups in this type of sampling technique is called clusters; and

b. stratified sampling requires that all strata must be represented while in cluster sampling, only selected clusters may be represented in the study.

There are also different types of cluster sampling:

(1) Single-stage cluster sampling. The researcher will select the clusters to be used through random sampling and then all the elements belonging in the chosen clusters must serve as subjects or respondents.

Step 1. If the target population are the residents of Zambales, it will be divided into towns (clusters).

Step 2. Random sampling will be done to identify which towns the researcher is going to use.

Step 3. After identifying, all citizens of the selected towns (clusters) will serve as sample.

(2) Two-stage or double-stage cluster sampling. The researcher will select the clusters to be used through random sampling and then another random sampling will be done within the cluster to select the sample. Step 1. If the target population are the residents of Zambales, it will be divided into towns (clusters).

Step 2. Random sampling will be done to identify which towns the researcher is going to use. However, not all residents in the identified towns will serve as respondents.

Step 3. Another random sampling will be done in each town to select the barangays that the researchers will use.

This means that unlike the single-stage cluster sampling, not all citizens in the selected towns will serve as sample but only those citizens living on the selected barangays.

(3) Multistage sampling in which the researcher, after doing the two-stage cluster sampling, will do another random sampling.

Step 1. If the target population are the residents of Zambales, it will be divided into towns (clusters).

Step 2. Random sampling will be done to identify which towns the researcher is going to use. However, not all residents in the identified towns will serve as respondents.

Step 3. Another random sampling will be done in each town to select the barangays that the researchers will use.

Step 4. To achieve the multistage sampling, not all citizens on each barangay will serve as sample as the researcher will randomly select citizens in the barangay to become participants.

We will notice that as we progress with the types of cluster sampling, sample size becomes smaller and smaller.

13 | P a g e

2.4.2. Non-probability Sampling. This is a type of sampling technique where the researcher’s subjective judgment is used to select the subjects.

Types of Non-probability Sampling

2.4.2.1. Convenience Sampling

• This is where participants included where chosen because they are conveniently available to the researcher.

• For example, a researcher conducting a study about vendors can only gather data after his classes has ended at 5 in the afternoon, therefore, only the vendors who are still at their posts at that time will be available for the study. Those who are selling in the early morning or in the evening will not be able to participate.

2.4.2.2. Snowball Sampling

• This is where participants are recruited to be a part of the study through other participants.

• For example, a researcher conducted a study about the farmers in San Antonio, Zambales. However, she was not able to have a complete list of farmers in the place. What she did was she interviewed a farmer that she knew and then ask him for a name of another farmer. The recruited farmer, after the interview, also referred other farmers. This was done until the researcher completed her desired sample size.

2.4.2.3. Purposive Sampling

• This is where the participants were chosen by the researcher using a set of criteria.

• For example, a researcher conducted a study about the perceptions of HUMSS graduates who will be taking up a medical course in college. Those HUMSS graduate who would not take up a medical course would not be included in the research study.

2.4.2.4. Quota Sampling

• This is like stratified random sampling. However, each group has an equal or proportionate representation of subjects.

• Groups maybe divided depending on the preferred variable of the researcher. It can either be age bracket, gender, grade level, etc.

• For example, a researcher conducted a study where junior high school students are the target population. Regardless of differences in number in each grade level, the researcher decided to get 25 participants each to reach the desired sample size of 100.

Determining the Sample Size

A. Margin of Error – it is the allowable error margin in research. Its main purpose is to identify how many percentage points the results will differ from the real population value.

B. Confidence Interval – it allows us to see the actual low and high limits of the estimate at a given significance level.

C. Confidence Level – it tells how confident the researcher is to the result of the study.