What was your independent variable? What question did it ask of a respondent? If you created your own dataset, what was the source of this data and did you cite it? How was your variable measured (continuous, categorical, binary, etc.) and did you transform or recode in any way? Did you remove any observations from your dataset? ‘
USEFUL NOTES FOR:
What was your independent variable? What question did it ask of a respondent? If you created your own dataset, what was the source of this data and did you cite it? How was your variable measured (continuous, categorical, binary, etc.) and did you transform or recode in any way? Did you remove any observations from your dataset?
Introduction
If you’re like me, then your data is probably messy and hard to understand. The goal of this guide is to help you make sense of what’s in your dataset by answering the following questions:
What was your independent variable?
Your independent variable is the variable that you are manipulating or testing for a relationship with other variables. It can also be referred to as your independent reasoning, because it is what you’re changing in order to see if there’s an effect on another variable.
In this case, it’s better to think of it as just another word for what we call “independent” when talking about statistics: something like “independent from.”
What question did it ask of a respondent?
The independent variable in your dataset is the question asked of a respondent. To determine what this question was, ask yourself:
What was the response to this question? For example, if you had an interviewer asking “How often do you buy groceries?” and someone responded with “3 times per week,” then they would be classified as a frequent grocery shopper.
How was it asked? Was there any preamble before asking them to respond (e.g., “What’s your favorite type of food,” followed by “How often do you buy groceries?”)? Or did they have an option to skip answering altogether (e.g., just write down their answer)?
If you created your own dataset, what was the source of this data and did you cite it?
If you created your own dataset, what was the source of this data and did you cite it?
If a dataset is not publicly available (e.g., from a government agency), then the researcher must provide information about where the data came from. If an investigator has access to proprietary databases or other restricted sources but does not wish to make them public, he or she should state that in their paper and cite their source if possible.
How was your variable measured (continuous, categorical, binary, etc.) and did you transform or recode in any way?
If you are using a continuous variable, then you need to transform it into a categorical variable.
If you are using a categorical variable, then you need to transform it into a continuous variable.
And if your dataset contains only two categories (binary), then it’s best not to include your data in this study at all and instead focus on analyzing other types of datasets that have more than two categories in them!
Did you remove any observations from your dataset?
As you go through the process of creating your own dataset, it’s important to be aware of what data elements can be removed from your dataset. In some cases, you may want to remove observations that don’t fit the research question or are outliers. For example, if a respondent had an extreme value and their behavior didn’t fit with other respondents in their group (or even within themselves), then removing them from consideration would improve overall validity for your study.
The takeaway is not to have a takeaway.
Oops! Click Regenerate Content below to try generating this section again.
Conclusion
In conclusion, if you are going to do data science, be sure that the conclusions of your analysis are clear and understandable. You will have more success with this if you keep things simple. Don’t try to make it too complicated or confusing; it’s okay if there aren’t any paragraphs in your report! But also don’t waste time on unnecessary details that won’t help others understand what you’ve done: think about how much information can be conveyed by each sentence without getting too wordy or verbose (e.g., “We found that…”).