If you’re confused and don’t know what to do if you have a research assignment in hand, this post is for you. Here is a guide on how to conduct a study and the different methods of doing so, depending on the situation.
Table of Contents
- Define your topic and method of data collection
- Conduct background research and predict the result
- Build or find materials for observation (if you need observations)
- Try obtaining existing data (if the data that will be used already exists)
- Interviews and surveys (if you want to collect opinions)
- Analyze data
- What will this mean to you?
- Finalize your ideas and start writing
- Cite your sources
- Proofread
- Include graphs
- Improve other aspects of your report
1. Define Your Topic and Method of Data Collection
When you conduct a research, make sure you define your topic clearly. If not, you cannot conveniently do the rest of the research. This includes finding how to collect data for the research, which includes:
- Environmental observations or experiments (refer to step 3)
- Existing records and data (refer to step 4)
- Interviews and Surveys (refer to step 5)
What will you choose? This heavily depends on what data you’re collecting. For instance, if you want to survey how certain chemicals interact with each other, you’ll probably have to do experiments. However, if you want to learn about the general math level of different large groups of people, you probably have to create online surveys.
However, you will want to make sure every aspect is covered and potential reasons are considered. For instance, if you want to find out why some operating systems may crash more often than others, you might have to produce the same type of problems and see how the different systems react. You need to test this on different computers to rule out any independent issue that only exists on a single computer.
Moreover, you can also make use of sampling. Finding a sample size that meets your requirements is crucial. What if you’re using the whole target audience? It might be a little costly and time-consuming, especially if you’re not using third-party tools or an API. But what if you’re only surveying a few people? This sample cannot accurately predict the results, as everyone knows.
2. Conduct Background Research and Predict the Result
Before dealing with the data, make sure you do background research to facilitate the study. This will give you some tips on how to proceed, including what to use to collect the data or attract more users to provide more data for the study. What you’ve found should enable you to write a detailed article about information related to the study (perhaps as long as this one). But these sources should be backed up with multiple sources and/or your reasoning skills!
Moreover, you should construct a hypothesis. This might give you room to write more in your report. However, make sure you think twice before proposing one, or there will be significant flaws if you include an unreasonable hypothesis in your report.
3. Build or Find Materials for Observation
If you have to use observations to obtain the results, you must conduct experiments first. In this case, you can make use of lab materials and find out how to produce the chemicals or materials needed for the solutions. You might also have to create programs to look for the answer. In some (probably) unlikely scenarios, you might even have to make use of engineering to build prototypes yourself.
Make sure you’ve noted the procedures acquired during background research, and pay serious attention to safety, and follow safety guidelines for specific chemicals during the process. You don’t want to obtain a serious injury from obtaining a simple answer. If you cannot remember or make up a list of procedures with certainty, try to search for more data before continuing.
4. Try Obtaining Existing Data
Sometimes, materials for your research are already accessible on some servers globally, but not so easily at the background research stage. Therefore, you will have to use more advanced methods to obtain data, such as using APIs. An API is a communication link between multiple pieces of software to exchange data and code, obtain permissions, and/or execute instructions on the other side.
Keep in mind that API security must be implemented seriously to prevent it from being attacked. This includes handling API keys appropriately. Therefore, you should keep API keys private and consider storing them in your back-end instead of in your code. Moreover, like changing passwords, regenerate your API key and deactivate the old one immediately if you suspect that your API key has been exposed to anyone else. Your team members probably don’t need your API key if you’re responsible for using the API, or you don’t need to care about this if you’re not the one in charge of that.
However, some researches don’t go this far and only involves searching for sources like books, newspapers, magazines, and webpages. If you need to collect data from large amounts of webpages, consider web scraping.
5. Interviews and Surveys
To collect opinions, you should use questionnaires, or if you want more extensive ones, interviews might be necessary. However, interviews might be more costly and time-consuming to conduct, so you probably don’t have to conduct interviews directly unless you have to use facial or speech recognition technology. For instance, if you want to detect whether liars have a facial expression that is different from truth-tellers, interviews do best.
Besides, you should purely rely on the opinion of the people who participate in the survey, and do not ask questions that seems to have a “better” answer. Failure to do so might result in inaccurate results after you’ve collected the data. Moreover, making the surveys interactive, like displaying branches of different questions after answering one, might help because the user wouldn’t need to scroll all the way down to the right spot, which might cause unnecessary confusion or even errors.
It’s time to define your objectives again, and this time, learn about what to know. You only need to observe the right thing, choose the proper API service, and/or ask necessary questions.
6. Analyze Data
Now, it’s time to analyze the data. The most convenient method is to use computers because you might make errors if you jot down a large number of data on paper.
For instance, you can use SQL to extract critical parts of the database. After that, you can use other tools, such as pandas in Python, to find important statistical information and even correlations. But what tools can you use exactly? This article is not teaching programming, but the methods must depend on what has to be obtained. For example, if you want to know whether certain types of content have a higher CTR than usual, you must see whether it’s a positive or negative correlation.
What is a positive (or negative) correlation? Take a look at the two variables representing the X and Y axis on diagrams. If, after one variable increases, the other also appear to go up, it’s a positive correlation. If a variable increases but the other goes down, it’s a negative correlation. Specifically, if the correlation coefficient is positive, it indicates a positive correlation. Otherwise, it’s a negative correlation.
7. What Will This Mean to You?
After results are out, it’s time to think about what that data means to you and the general population. You might need more data to evaluate how to improve lives based on this data. Consult professional advice if necessary.
For instance, if data indicates that a flawed piece of code in a certain system causes it to crash more often, try persuading users to access the associated feature only if necessary, save all unsaved data before doing so in case the system crashes, and regularly check for updates to see if this bug is fixed. In this case, you might also need to follow up and update the paper if the issue is fixed after the initial publication. If the problem is already fixed at the time you publish it, mention that the issue has fortunately disappeared, and try more general precautions, like saving unsaved data regularly.
However, suppose this issue is persistent and takes a long time to resolve, like (hopefully rare) genetic issues that will pass on through generations. In that case, you can confidently write prevention tips without the “error recovery code” above, but make sure you get opinions on professionals somehow for these relatively unpredictable problems if necessary and mention them in your report.
8. Finalize Your Ideas and Start Writing
At this point, you’ve probably acquired sufficient information for you to write the report. In that case, you can note down your ideas and start writing. It’s like writing assignments you’ve done before, so you shouldn’t have trouble writing your essay. However, after writing your essay, it’s crucial that you proofread and optimize it to give your readers a great impression.
9. Cite Your Sources
However, before you continue, make sure you’ve cited all your sources (i.e. that your bibliography is complete and that it’s fully listed in the essay). You might face severe consequences for failing to do so!
Before or after you interview a person who gives you professional opinions on things to include in the essay, you should list them. However, if you’ve forgotten to cite one or more webpages, try searching in the browsing history. Make sure you comply with citation formats anyway. You might need to use a specific format, depending on your assignment.
10. Proofread
After you’ve finished writing your article, you have to ensure that it’s free of grammatical and spelling mistakes. If you don’t, you might silently impair user experience by showing that you’re not proficient in the language you’re using on the report! Besides avoiding errors in the first place, it’s also important to use online proofreading apps like Grammarly in case you missed anything.
Besides, you should see if your sentences are too wordy, or if the messages are too indirect or well-hidden. These clarity issues impair user experience and should be resolved, too.
Although on-page SEO techniques should be implemented, don’t overdo it. Keep the flow and content of the article natural, or the readers will not like it. For example, who wants a 4-word phrase to have more than 20 occurrences in a 1000-word article? It will look not very interesting and possibly impact the quality! Instead, only add keywords if necessary, and make sure the topic is clear and matches what was given in the assignment.
11. Include Graphs
In a report, it’s important to include graphs. If you get to a webpage and see thousands of lines of raw data, you’ll likely find it difficult to read (or even fishy) and find another relevant page. As a result, instead of the data itself, you should use graphs to represent it in a more human-friendly way.
Whether they’re pie charts, bar charts, histograms, or any other types of charts, they help visualize data to make your paper easier to read. Other than retaining users, it can offer many other benefits, such as helping you promote your report on social media.
However, you must not make graphs misleading. If they are, it will provide the readers with incorrect information if they don’t look at the graphs carefully. As such, if you want to exaggerate the differences, think again. If this is not quite significant after all, you should not zoom in on just one part of the graph.
12. Improve Other Aspects of Your Report
It’s essential that you improve other aspects of your report for your readers as well. For instance, you should:
- Break up paragraphs and include subheadings
- Make headlines compelling, but not clickbait
- Make a glossary for less-known words
- Include external and internal links in the right places (such as linking to the websites in the bibliography)
- And more
Conclusion
These are the things that you should do in a research, depending on the situation. It might seem tough at first, but after you know what to do, you will obviously feel more determined than otherwise. You can always make up more tips or steps, but make sure you put them in the comments below for us to improve the article. For more details on the tips, please read the webpages from the references below.
References and Credits
- (2019, July 23). 7 Data Collection Methods & Tools For Research. Retrieved September 1, 2021, from https://www.formpl.us/blog/data-collection-method
- (2019, September 27). Data Collection Methods. Retrieved September 1, 2021, from https://www.jotform.com/data-collection-methods/
- Brown, J. L. (2017, April 18). The Essential Guide to Writing Effective Survey Questions. Retrieved September 2, 2021, from https://www.uxbooth.com/articles/the-essential-guide-to-writing-effective-survey-questions/
- Stephanie Briggs. (2015, June 25). Surveys 101: A Simple Guide to Asking Effective Questions. Retrieved September 2, 2021, from https://zapier.com/learn/forms-surveys/writing-effective-survey/
- (n.d.). Using API keys. Retrieved September 2, 2021, from https://cloud.google.com/docs/authentication/api-keys
- Pincovai, M. (2020, July 17). 5 best practices for secure API key storage. Retrieved September 2, 2021, from https://developers.amadeus.com/blog/best-practices-api-key-storage
- Investopedia, Drury, A. (n.d.). What Does a Negative Correlation Coefficient Mean? Retrieved September 2, 2021, from https://www.investopedia.com/ask/answers/041015/what-does-negative-correlation-coefficient-mean.asp
- Nickolas, S. , Mansa, J. (n.d.). What Do Correlation Coefficients Positive, Negative, and Zero Mean? Retrieved September 2, 2021, from https://www.investopedia.com/ask/answers/032515/what-does-it-mean-if-correlation-coefficient-positive-negative-or-zero.asp
- Greene J. (2019, June 24). SEO Copywriting Tips for Writing High-Quality, High-Ranking Content. Retrieved September 3, 2021, from https://databox.com/seo-copywriting-tips