91Ó°ÊÓ

Non-response bias occurs when there is a significant difference between those who respond to a survey and those who do not. When only certain groups within the population complete a survey, the findings might not accurately represent the whole population's views or behaviors. This is vital in survey research because non-respondents may have different opinions or ways of using the Internet than those who responded, perhaps due to different education levels affecting their likelihood of answering.
For instance:

• If individuals with lower levels of education are less likely to participate in the survey, possibly due to lacking interest or access, then the results might incorrectly suggest that internet use habits differ significantly only among those with higher education.
• Conversely, if those with higher education levels are more inclined to answer the survey, it may exaggerate their internet usage patterns relative to those less educated.

In conclusion, non-response bias can significantly distort survey results, making it crucial for researchers to address this issue through techniques such as follow-up reminders or incentives for participation.

Selection bias emerges when the sample used in a survey is not representative of the larger population, often due to how participants are selected. In the context of the exercise, if the 1300 households were not chosen randomly or do not encapsulate the diversity present in the entire country, the study's conclusions might be skewed.
Some common causes of selection bias include:

• Geographic or demographic overrepresentation or underrepresentation, which means certain areas or groups might be omitted or overly included.
• Using convenience sampling rather than random sampling, leading to certain biases based on the survey's accessibility.

Addressing selection bias involves ensuring that participants are randomly selected and adequately represent the demographics of the broader population. This helps to generate findings that are realistically applicable and dependable.

Survey methodology encompasses the overall process of designing and conducting surveys to collect reliable data. The methodology should be robust to mitigate any biases and ensure valid conclusions. A well-thought-out survey methodology includes several components such as the design of the questionnaire, sampling method, and the data collection process.
Key aspects to consider:

• The mode of survey distribution, like telephone, online, or mail, affects who is more likely to respond. For example, digital surveys may miss audiences with limited internet access.
• Choosing the correct sample size is crucial. It should be large enough to be statistically significant while considering logistical constraints.
• Ensuring anonymity and confidentiality to encourage honest and accurate responses from participants.
• Implementing follow-up techniques like reminders to improve response rates and reduce potential non-response bias.

By carefully crafting survey methodology, researchers can minimize bias risks and maximize the integrity and generalizability of the survey's conclusions.