Quant Skills for Qual researchers
Take your career to the next level by broadening your skills
Are you looking to enhance your quantitative research abilities but feeling overwhelmed by the plethora of options available? Are you interested in becoming a mixed-methods researcher but unsure of where to begin?
You are not alone.
There are many resources available for learning more about quantitative skills, but it can be difficult to know which ones are the best starting points. In this blog, I aim to address this issue and provide guidance for those looking to improve their quantitative research abilities.
Here are the most important skills you need to learn to become a mixed-method researcher:
Familiarize yourself with dashboards
Build survey skills
Brush up on basic statistics
Learn a tool for analysis and visualization
Let’s dive in.
1- Familiarize yourself with dashboards
When looking to improve your quantitative research abilities, it's important to first familiarize yourself with the dashboards and other resources that are already available. These tools can provide valuable insights and save you time by eliminating the need for data collection and extensive analysis. Keep in mind that understanding how dashboards work and interpreting the underlying data can be time-consuming. Don't hesitate to reach out to experts such as data scientists, engineers, or quantitative UX researchers for assistance in navigating these resources. The best way to improve your skills is by learning through hands-on experience and seeking guidance from those who have more experience in the field.
It's also important to remember to fully understand the data, metrics, and charts before using them in your reports or relying on them as insights. Be sure to ask questions and seek clarification if you have any doubts. This will ensure that your interpretations are accurate, reliable, and actionable.
2- Build survey skills
Surveys are a valuable tool for collecting data from a specific population and gaining insights into demographics and attitudes. They can be administered in various ways, such as online, over the phone, or in-person, but online surveys are often the most convenient option.
Designing surveys and running them may sound easy but there is a lot of science behind it that if overlooked could risk the accuracy of your results. But it is also easy to become overwhelmed with the various skills and knowledge required to run a successful survey. To streamline the process, I suggest you mainly focus your energy in learning good questionnaire design and understanding survey sampling.
Learn how to design good questionnaires
Poor questionnaire design can compromise the validity of survey results. Some common issues that can arise from poor questionnaire design include:
Ambiguity: Questions that are poorly worded or unclear can lead to confusion and misinterpretation, resulting in inaccurate or unreliable data.
Leading questions: Biased or suggestive questions can influence participants to provide responses that may not be accurate or representative of their true beliefs or behaviors.
Double-barrelled questions: Asking multiple questions within one question can lead to confusion and make it difficult to obtain clear and specific answers.
Lack of response options: Not providing enough response options or not including an "I don't know" or "other" option can limit participants' ability to provide a full range of answers.
Order effect: The order in which questions are presented can affect the responses received. For example, if a question about a sensitive topic is asked before a question about a less sensitive topic, it could lead to a higher rate of non-response.
Sensitivity: Asking sensitive questions in the wrong way could lead to respondents feeling uncomfortable and not responding truthfully.
Pew Research has a great write up and a video on writing survey questions. Another good starter is this old, but still relevant tip sheet on survey question wording.
Take sampling very seriously
Proper sampling is crucial to ensuring accurate survey results. Sampling can be affected by various types of biases and errors, making it essential to use appropriate sampling techniques.
Common issues with survey sampling include:
Selection bias: This occurs when the sample is not representative of the population being studied because you only targeted survey to a certain segment or group of respondents.
Nonresponse bias: This occurs when certain groups of people are less likely to respond to the survey, leading to a sample that is not representative of the population.
Response bias: This occurs when the way the survey is designed or administered leads to inaccurate responses. Leading questions or biased wording can lead to inaccurate responses.
Sampling error: This occurs when the sample is not a perfect representation of the population.
Self-selection bias: This occurs when individuals choose to participate in the survey on their own, rather than being randomly selected. This can lead to a sample that is not representative of the population.
Here is a quick read on the topic of survey sampling. For a deeper understanding of survey sampling, check out the book, Survey Sampling by Leslie Kish.
3- Brush up on basic statistics
Generally you need a solid understanding of probability and statistics to start on any data analysis project. Here are the most essential ones to review:
Descriptive statistics: It involves summarizing and describing the data using measures such as mean, median, mode, standard deviation, and variance. These measures provide a quick overview of the data and allow for identifying patterns and trends.
Probability distributions: Probability distributions are used to model random events. Common probability distributions include normal, binomial, and Poisson.
Hypothesis testing: It is a statistical method used to test a claim or hypothesis about a population parameter. Some common hypothesis tests are t-test and chi-squared test.
Inferential statistics: It allows for making inferences about a population based on a sample of data. This is done by using probability and statistical theory to estimate population parameters, such as means and standard deviations. Inferential statistics also include methods for constructing confidence intervals and for comparing means and proportions between groups.
There are lots of resources on statistics online, here is a free course on beginner statistics.
4- Learn a tool for analysis and visualization
Finally, to effectively analyze data, you need the following skills:
Data Wrangling: The ability to effectively manipulate, clean, aggregate, and reshape data is crucial for preparing data for analysis.
Data Visualization: The ability to create clear, informative visualizations that represent and communicate data is important for understanding and communicating the results of your research. For example, knowing when to use histograms, scatter plots, etc. and how to create them from your data.
Statistical Programming: The ability to run statistical tests using tools such as R or Python is also necessary. There are many libraries available for statistical analysis in R and Python, making it relatively easy to use these tools for this purpose.
Here is a great free course on statistics with R.
To perform analysis, it is essential to be familiar with a few data analysis tools. While many people may feel that learning a statistical programming language such as R or Python is necessary, it is not the only starting point. For some of your analysis needs you could even use spreadsheets.
List of resources for further learning:
Improving Survey Questions: Design and Evaluation, Floyd J. Fowler
Asking questions: A practical guide to questionnaire design, Seymour Sudman and Norman Bradburn
Survey Sampling, Leslie Kish