Quantitative Data Collection
Topic 2: Data Collection
Topic 2: Data Collection
The choice of data collection method is a critical point in the research process. Quantitative data collection typically involves one or more of the following:
A clear and concise research question is essential before determining how you will collect your data. Moreover, the type of data you need to answer your research question (e.g., performance or achievement levels, attitudes or opinions, knowledge or skills, etc.) along with the complexity of the information needed should be considered prior to choosing the appropriate data collection strategy. Finally, be sure to consider the feasibility of each method and the availability of participants to ensure that you will be able to collect the data you need.
Other factors involved in determining your method include the following:
- Surveys, tests, or questionnaires – administered in groups, one-on-one, by mail, or online;
- Reviews of records or documents using a rubric; or
- Observations.
A clear and concise research question is essential before determining how you will collect your data. Moreover, the type of data you need to answer your research question (e.g., performance or achievement levels, attitudes or opinions, knowledge or skills, etc.) along with the complexity of the information needed should be considered prior to choosing the appropriate data collection strategy. Finally, be sure to consider the feasibility of each method and the availability of participants to ensure that you will be able to collect the data you need.
Other factors involved in determining your method include the following:
- Accuracy/trustworthiness of data collected (i.e., respondent can openly answer; limited researcher bias)
- Amount of data that can be collected easily
- Degree of sampling bias (i.e., ability to draw a representative sample)
- Cost and effort required
- Speed (i.e., time it takes to complete data collection)
- Administrative issues (i.e., management, auditing, data entry)
- Depth of information that can be obtained
| Factors | Surveys/Tests/Questionnaires | Record Review w/ Rubric | Observation | |||
| Group administration | 1:1 interview* | Mailed survey | Online survey* | |||
| Accuracy of data collected | Very good | Fair to very good | Very good | Very good | Fair to very good | Fair to very good |
| Amount of data that can be collected | Limited | Limited to extensive | Limited | Limited to extensive | Limited to extensive | Extensive |
| Degree of sample bias | Moderate | Moderate | Moderate to high | Moderate to high | Moderate to Low | Moderate to Low |
| Cost and effort required | Moderate | Moderate to high | Low | Very low | Moderate | Moderate |
| Speed (real time) | Fast | Fast | Slow | Very fast | Fast | Fast |
| Administrative issues | Start-up and Data entry | Constant | Constant | Start-up | Constant | Constant |
| Depth of information | Low to moderate | Low to moderate | Low | Low to moderate | Moderate | Moderate |
| *Online surveys and one-on-one interviews allow for more in-depth information to be collected than paper mailed surveys and group administrations because you can build in extensive skip patterns and complex questions; paper surveys should refrain from complex question structures or skip patterns that might confuse participants. | ||||||
Once you have determined the best method for your research question, you will need to either locate existing (based on a literature review) or create new instruments (e.g., surveys, rubrics, observation checklists) to collect the data. It is recommended that you use existing tools that have already been tested in the field and validated whenever possible. Should you need to construct your own instrument, please refer to the course materials from your Inquiry courses and some of the additional resources listed below.
Also be sure to pilot test all instruments before you use them in your research.
Prior to collecting your data, you will also need to determine if you will be collecting data from everyone in your population of interest (a census) or if you will be collecting data from a sample. There are two types of sampling to consider: probability sampling (random sampling: everyone in the population has an equal opportunity to be selected for the research) or non-probability sampling (convenience or purposeful sampling). Determining the appropriate size of your sample depends on the nature of your study (descriptive, correlational, causal) and your research design. If you are going to use a sample, you should consult the resources below to learn more about sampling and its implications for your study.
Finally, be sure to standardize your data collection (i.e., administer surveys the same way each time; conduct interviews with the same procedures; track data consistently) in order to maximize the reliability and accuracy of your data.
Also be sure to pilot test all instruments before you use them in your research.
Prior to collecting your data, you will also need to determine if you will be collecting data from everyone in your population of interest (a census) or if you will be collecting data from a sample. There are two types of sampling to consider: probability sampling (random sampling: everyone in the population has an equal opportunity to be selected for the research) or non-probability sampling (convenience or purposeful sampling). Determining the appropriate size of your sample depends on the nature of your study (descriptive, correlational, causal) and your research design. If you are going to use a sample, you should consult the resources below to learn more about sampling and its implications for your study.
Finally, be sure to standardize your data collection (i.e., administer surveys the same way each time; conduct interviews with the same procedures; track data consistently) in order to maximize the reliability and accuracy of your data.
 |
 |
