How does a researcher address response rate, missing data, and the errors they create?

Low response rates are a recurring problem in data collection methods such as surveys, where the respondents are not obliged to respond, perhaps because they are too shy, or too busy to do so. The failure by survey participants to respond may result in a type of error called non-response bias. Non-response error occurs when there is a clear difference between those participants who respond and those who do not (this must not be confused with another type of error called response bias, which occurs when there is a clear difference between the responses of the participants and their actual values, maybe due to inaccurate responses or poor record of responses). A researcher can address the problem of low response rates by employing any one of the following methods:

Thoroughly test the performance of the survey medium before conducting the survey, so as to ensure a smooth experience for any participant. If the survey is run online, for instance, it is important to ensure that it can be accessed from various devices (desktop, mobile, tablet) and that its design is user-friendly (readable font, nice use of colors, etc.)
Make sure the survey is not too long, tiring the respondent or requiring too much time to fill
Ensure confidentiality
Make the survey interesting, so that participants are encouraged to complete it
Say thank you to the participant, for taking his or her time to complete the survey

Missing data occurs when a data set has missing data values for one or more variables. This can be because of non-response of participants, or study participants dropping out in the course of the study, or even data entry problems. Missing data can be a big problem during data analysis as most statistical tools require complete data values for all given variables in order to complete computations. Also, when a variable has many missing values, statistics drawn from it may be less accurate in estimating parameters of the general population. Hyun Kang, in his article titled “The Prevention and Handling of Missing Data” gives the following methods for dealing with missing data:

Proper planning of the study to eliminate the problem of missing data values
Proper training of personnel on data entry methods
Set targets on unacceptable levels of missing data
To combat missing data during data analysis, use of robust analysis methods that ensure little bias in inferences made on the population
Use of case deletion—the omission of all cases with missing data in the final analysis
Use of Pairwise deletion—omission only of particular missing data points
Mean substitution—substituting missing data points with the means of the specific variables, etc.
https://business.ucf.edu/wp-content/uploads/2014/11/How-Low-Should-You-Go..Low-Response-Rates-and-the-Validity-of-Inference-in-IS-Questionnaire-Research.pdf