SAHAN Research and Consultancy

Overview of Research Introduction part ( Chapter one)
Follow for more.

MEAL: ( monitoring, evaluation, accountability, and learning)

Techniques
Sampling techniques are methods used to select a subset (sample) from a larger group (population) to draw conclusions or make inferences about the entire group. They are broadly classified into two main categories: probability sampling and non-probability sampling.
1. Probability Sampling: Every member of the population has a known, non-zero chance of being selected. It's more statistically rigorous.

- Simple Random Sampling: Every individual has an equal chance of being selected.
- Systematic Sampling: Every kth item is selected from a list after a random start.
- Stratified Sampling: Population is divided into strata (groups) and random samples are taken from each group.
- Cluster Sampling: Population is divided into clusters, some clusters are randomly selected, and all individuals in selected clusters are studied.
- Multistage Sampling: A combination of the above techniques used in stages.

2. Non-Probability Sampling: Not all individuals have a known or equal chance of being selected. It's often used in exploratory research or when probability sampling is impractical.

- Convenience Sampling: Selection based on ease of access.
- Judgmental/Purposive Sampling: Researcher selects individuals based on their judgment.
- Quota Sampling: Population is segmented into groups and a specific number is taken from each.
- Snowball Sampling: Existing participants recruit future participants from among their acquaintances.

Methods of data Collection

SAHAN R and C provide the following services.

Models:
Regression models are statistical models used to estimate the relationship between a dependent variable and one or more independent variables. They are widely used in various fields of study, including economics, finance, and social sciences, to analyze and predict the behavior of a dependent variable based on the values of one or more independent variables.

There are several types of regression models used in research and analysis. They are:

1. Linear regression model: The linear regression model is a basic regression model that assumes a linear relationship between the dependent variable and one or more independent variables. It is most commonly used in research and analysis and is a simple, yet powerful tool for modeling relationships between variables.

2. Multiple regression model: The multiple regression model is similar to the linear regression model, but it allows for the analysis of more than one independent variable. It is often used in studies that involve multiple variables that can affect the dependent variable.

3. Polynomial regression model: The polynomial regression model is used when the relationship between the dependent variable and the independent variable is curvilinear, rather than linear. It is a nonlinear regression model that fits a polynomial equation to the data.

4. Logistic regression model: The logistic regression model is used when the dependent variable is binary or dichotomous - that is, it can take on one of two possible values. It is used to predict the probability that a given event will occur based on the values of one or more independent variables.

5. Time series regression model: The time series regression model is used when the dependent variable is a time series variable and the independent variables are other time series variables. It is used to analyze and predict trends and patterns in time series data.

In summary, the type of regression model used for analysis depends on the research question and the nature of the data.

By selecting the appropriate regression model, researchers can analyze and understand the relationships between variables and make informed predictions about their behavior.

Brief Article Review Guideline
1. Citation: write the following information about the article you are going to review
👉Include full title, all authors (last name, initials), full journal title, year, volume number, and page numbers.

2. Summary:
👉 Present summary of essential contents and main ideas

3. Study Titles, problem &purpose
👉 is the title important? Self explanatory?
👉 Was the purpose and/or research question stated clearly?
👉 A clear statement of purpose or research questions helps you determine if the topic is important, relevant, and of interest to you.
👉 Does the author clearly define a research problem or topic? 👉 is its significance explained? Are core issues or research variables identified?
👉 Is specialized terminology usefully defined?

4. Was relevant background literature reviewed?
👉 A review of the literature should be included in an article describing research to provide some background to the study.
👉 Does the author provide an adequate literature review?
👉 Does it discuss current research on the problem, and help to situate the authors own research?

5. Are the research objectives clearly stated?
👉 Are hypotheses or specific research questions identified?

6. Methodology
👉 Does the author clearly identify the research methodology and any associated limitations of the research design?
👉 Are participants described, including the method of sample selection if appropriate?
👉 Are instruments adequately described, including issues of appropriateness, validity and reliability?
👉 Do any evident biases or ethical considerations arise in relation to the methodology?
👉 Are the methods for measuring results clearly explained and appropriate?

7. Results
👉 What are the author's major findings and conclusions?
👉 Have these been supported by the author's analyses, arguments, findings or evidence?
👉 Has the author overlooked anything?

8. Discussion & conclusion
👉 Do the research results validate the authors conclusions and/or recommendations?
👉 is the finding adquetly discussed and the author compared the finding with other similar studies??
👉 is the conclusion reported inline with its objectives??
9. Suggestion for future research:
👉 Does the author suggest areas for further research or discussion?

10. References
👉 Are references given (footnotes or bibliography)?
👉 What is the size of the reference section?
👉 Are the references recent, important? If you want any support or assistance in your studies related to semester projects, term papers, case studies, proposal

Classical linear regression relies on several key assumptions to ensure that the estimates of the regression coefficients are valid and that the inferences drawn from the model are reliable. Violating these assumptions can lead to biased or inefficient estimates. Here are the main assumptions and their implications when violated:

1. Linearity: The relationship between the independent variables and the dependent variable is linear.

• Violation Implications: If the true relationship is nonlinear, the model will misestimate the effect of the predictors, leading to biased coefficients and poor predictions. Nonlinear patterns may require transformations or polynomial terms.

2. Independence: The observations are independent of each other.

• Violation Implications: If there is correlation between observations (e.g., in time series data), this can lead to underestimated standard errors, resulting in misleading significance tests and confidence intervals.

3. Homoscedasticity: The variance of the errors is constant across all levels of the independent variables.

• Violation Implications: If the variance of the errors changes (heteroscedasticity), it can lead to inefficient estimates and biased standard errors, affecting hypothesis tests and confidence intervals.

4. Normality of Errors: The errors (residuals) are normally distributed.

• Violation Implications: While this assumption is not critical for estimating coefficients, it is important for conducting hypothesis tests and constructing confidence intervals. Non-normality can lead to unreliable p-values, especially in small samples.

5. No Perfect Multicollinearity: The independent variables are not perfectly correlated with each other.

• Violation Implications: Perfect multicollinearity makes it impossible to estimate the coefficients uniquely, leading to inflated standard errors and unreliable coefficient estimates. Even near-multicollinearity can cause issues with interpretation and stability of the estimates.

6. No Autocorrelation: In time series data, the residuals should not be correlated with each other.

• Violation Implications: Autocorrelation can indicate that a model is misspecified (e.g., missing a relevant variable or using an incorrect functional form). It leads to inefficient estimates and biased standard errors, impacting hypothesis tests.

7. Exogeneity: The independent variables are not correlated with the error term.

• Violation Implications: If this assumption is violated (endogeneity), it typically results from omitted variable bias, measurement error, or simultaneous causality. This leads to biased and inconsistent coefficient estimates.

▎Addressing Violations

When these assumptions are violated, there are various methods to address the issues:

• Transformations (e.g., log transformations) can help with linearity and homoscedasticity.

• Robust Standard Errors can be used to correct for heteroscedasticity.

• Adding Variables or using instrumental variables can help with endogeneity.

• Generalized Least Squares (GLS) or other modeling techniques can be employed to handle autocorrelation.

In practice, it's essential to conduct diagnostic tests (like residual plots, variance inflation factor for multicollinearity, etc.) to check for these violations before relying on the results of a linear regression analysis.

The picture of the book is available in the telegram channel.
=====
https://t.me/SAHANRC

✓✓ Definition of Type I Error

In statistics, type I error is defined as an error that occurs when the sample results cause the rejection of the null hypothesis, in spite of the fact that it is true. In simple terms, the error of agreeing to the alternative hypothesis, when the results can be ascribed to chance.

Also known as the alpha error, it leads the researcher to infer that there is a variation between two observances when they are identical. The likelihood of type I error, is equal to the level of significance, that the researcher sets for his test. Here the level of significance refers to the chances of making type I error.

E.g. Suppose on the basis of data, the research team of a firm concluded that more than 50% of the total customers like the new service started by the company, which is, in fact, less than 50%.

✓✓ Definition of Type II Error

When on the basis of data, the null hypothesis is accepted, when it is actually false, then this kind of error is known as Type II Error. It arises when the researcher fails to deny the false null hypothesis. It is denoted by Greek letter ‘beta (β)’ and often known as beta error.

Type II error is the failure of the researcher in agreeing to an alternative hypothesis, although it is true. It validates a proposition; that ought to be refused. The researcher concludes that the two observances are identical when in fact they are not.

The likelihood of making such error is analogous to the power of the test. Here, the power of test alludes to the probability of rejecting of the null hypothesis, which is false and needs to be rejected. As the sample size increases, the power of test also increases, that results in the reduction in risk of making type II error.

E.g. Suppose on the basis of sample results, the research team of an organisation claims that less than 50% of the total customers like the new service started by the company, which is, in fact, greater than 50%.

✓✓ Key Differences Between Type I and Type II Error

1. Type I error is an error that takes place when the outcome is a rejection of null hypothesis which is, in fact, true. Type II error occurs when the sample results in the acceptance of null hypothesis, which is actually false.

2. Type I error or otherwise known as false positives, in essence, the positive result is equivalent to the refusal of the null hypothesis. In contrast, Type II error is also known as false negatives, i.e. negative result, leads to the acceptance of the null hypothesis.

3. When the null hypothesis is true but mistakenly rejected, it is type I error. As against this, when the null hypothesis is false but erroneously accepted, it is type II error.

4. Type I error tends to assert something that is not really present, i.e. it is a false hit. On the contrary, type II error fails in identifying something, that is present, i.e. it is a miss.

5. The probability of committing type I error is the sample as the level of significance. Conversely, the likelihood of committing type II error is same as the power of the test.



page