Thanksgiving Journey Through Data : Day 1

Holidays and traditions are more than annual events; they offer a glimpse into the heart of culture, community, and individual values. Major holidays like Thanksgiving provide a wealth of data that reflects trends in social norms, regional habits, and economic factors. With the use of an advanced data platform like Querri, that allows users to pose straightforward questions to complex datasets, we can uncover insights that were once obscured by technical hurdles. Querri is thrilled to take our readers on a **5-day Thanksgiving journey Through Data**. We'll travel back in time to uncover what was on America’s plate that Thanksgiving night in 2015. Here is a list of what we will cover over the next 5 days: - **Day 1**: Exploration of the Survey participants —who they are, where they’re from, and how they are distributed throughout the country.** - Day 2**: Sweet tooth alert! 🍰 Dive into America’s dessert favorites and regional pie debates.** - Day 3:** Sides steal the show. 🥗 From **stuffing** to green bean casserole, we’ll uncover the sides that bring families together.** - Day 4:** The main event: **Turkey**** vs. alternatives**—what was trending and who was replacing turkey with other alternatives.** - Day 5:** Beyond the table—**Black Friday** shopping trends and the growing popularity of **Friendsgiving **a decade ago ## The Dataset This dataset is mainly about the food on the table, but there is also great information about people. 1058 people responded to the [538 Thanksgiving Dinner survey](https://github.com/fivethirtyeight/data/tree/master/thanksgiving-2015) conducted in 2015. They were from all over the country. The list of survey questions can be viewed at the link above. ## Kicking Things Off With a Cleanup As we all know, high-quality data is the cornerstone of meaningful insights. That’s why we started by giving the data frame a thorough cleanup. We asked Querri to clean the data for us with a simple prompt, "***Clean the data frame***," and viola, it cleaned the headers, formatted the data for consistency, prompted us to handle the missing values, and categorically organized it into age and income groups. **Cleaning data with Querri's Cleaner tool was a breeze!** ## Exploratory Data Analysis (EDA) Next, as any expert data scientist would, we performed a quick exploratory data analysis (EDA) to get a clear picture of our dataset before diving deeper. But mind you, we are no data scientists, yet we could pretend to be one because Querri could do all the work for us (sh, 🤐 that's our little secret) with just one simple command. Exploring data is not just about numbers; It helps teams: - **Spot Patterns**: Uncover trends that shape smarter strategies and better products. - **Segment Smartly**: Classify data into actionable categories for enhanced efficiency. - **Stay Agile**: Quickly adapt to shifts using easy-to-use tools. We performed the EDA by typing a three-letter command "*EDA*". And in seconds, it delivered a concise summary of the dataset, key observations and even suggested next steps. ** * {% icon icon_set="fontawesome-6.4.2" name="Rocket.Chat" height="25" purpose="decorative" title="Rocket.Chat icon" %} `Querri Prompt: `**`EDA`* Click to zoom ## ## Suggestions We are thankful for Suggestions in Querri! It is such a great tool - almost like your mentor who guides you when you are stuck or your thinking partner when you need more than one brain to think. One of the suggestions that we accepted from the EDA step above was to visualize the participants on the US map. And just like that, with a single click to accept the suggestion, Querri laid out the entire survey population on the interactive US map using its **Geospatial** tool. {% icon icon_set="fontawesome-6.4.2" name="Rocket.Chat" height="25" purpose="decorative" title="Rocket.Chat icon" %} Querri Prompt: Visualize the participants on a US map Click to zoom The map above displays the distribution of respondents across various U.S. states, with each state shaded according to the number of respondents. Darker shades of blue indicate a higher number of respondents, while lighter shades show fewer participants. The most amazing thing about this visualization is that the dataset did not contain any states, just the regions, and yet Querri could lay out the respondents on the map accurately. The bar chart below shows another form of visualization of respondents in each region. | `***{% icon icon_set="fontawesome-6.4.2" name="Rocket.Chat" height="25" purpose="decorative" title="Rocket.Chat icon" %} Querri Promp**t: Show a graph of the percentage of respondents in each region*`| | | [](https://cms.qtm.querri.comhttps://media.querri.com/respondents_20by_20region_20_201_ddf54c30c0.jpg)| | | `***{% icon icon_set="fontawesome-6.4.2" name="Rocket.Chat" height="25" purpose="decorative" title="Rocket.Chat icon" %} Querri Promp**t: Make it colorful*`| | | | | Click to zoom ## No State? No Problem! Querri Saves {% icon icon_set="fontawesome-6.4.2" name="Clock" height="32" purpose="decorative" title="Clock icon" %} As we mentioned, the dataset contained only regions, not States in each region. Traditionally, in Excel, we would google each of the regions and spend at least 45-60 minutes searching, populating the 18 regions, and copy-pasting across all the 1000 rows, but Querri saved the hour and was able to populate all the states for the region in **l****ess than 5 minutes**. While we didn’t quite get Querri to populate all the regions on the first try, it was able to do so in 2 more attempts. | | | | [](https://cms.qtm.querri.comhttps://media.querri.com/States_20_Prompt_1_39e555a6cb.jpg)| | Click to zoom ## Trend Analysis by Age and Gender We wanted to examine the age distribution of respondents across the region, so we asked Querri to perform a trend analysis on age.Querri responded by a) creating a pivot table by US region and age, b) performing a statistical trend analysis on age distribution, and c) creating a stacked bar chart of age distribution in each region. One prompt, 3 outputs, zero formulas/lines of code! Querri is like your over-zealous friend who goes above and beyond every time you ask them for a favor because they just love you so much 😀 | *** {% icon icon_set="fontawesome-6.4.2" name="Rocket.Chat" height="25" purpose="decorative" title="Rocket.Chat icon" %} `Querri Prompt: `***`*Conduct a trend analysis of the age distribution across different us-regions*`| | | [](https://cms.qtm.querri.comhttps://media.querri.com/Trend_20analysis_20_201_1_53ecd1108e.jpg)| | | | | | [](https://cms.qtm.querri.comhttps://media.querri.com/Trend_20_Analysis_20_202_c52814dccb.jpg)| | Click to zoom I did the same for gender, though we could have combined all of this into one prompt. | ***{% icon icon_set="fontawesome-6.4.2" name="Rocket.Chat" height="25" purpose="decorative" title="Rocket.Chat icon" %} ***`***Querri Prompt**: Visualize gender by region*`| | | | | Click to zoom As you can see, you can keep your prompts short and straightforward for standard data tasks—Querri will still understand you! (Though for more complex tasks, more specificity might be needed.) ## Practicing Good Judgment with Human Data Ok, by now, we have a fairly good idea of how the respondents are distributed across the country and their demographics - age and gender. You may be thinking, but what about the income levels? What demographic data about adult human beings is complete without income information? And even though this dataset contained income levels of the respondents in each region, we are holding back on sharing the income level information to highlight something important: **behind every data point about people are real human lives**. It’s easy to get caught up in the numbers and forget the human stories they represent. Data doesn’t make decisions—we do. That’s why it’s so critical to use caution and good judgment when analyzing or sharing data, especially when those decisions could impact others. Treat data about people with the care and responsibility it deserves. ## Calling It a Day on Day 1 With our initial exploratory analysis of respondents, we conclude our day 1. Today, we: a. Cleaned the data using the Cleaner tool b. Performed exploratory data analysis using the Geospatial tool c. Visualized distribution of respondents on the US map d. Added a new column for US states and populated it based on each region d. Performed a trend analysis and visualized respondents by age and gender ## In Conclusion... Just like families gathering to organize the perfect dinner, companies of all sizes are gathering their data, piecing it together for insights to guide them forward. Business, after all, is a bit like Thanksgiving dinner—rooted in tradition, connection, and decisions that bring people together. In the spirit of Thanksgiving, we at Querri are thankful for the opportunity to help businesses harness the power of data. Whether you're diving into holiday trends or seeking everyday business insights, Querri is here to provide the tools you need to make data-driven decisions. Let's celebrate growth, insights, and the spirit of gratitude together! *From mountains high to valleys low* *Through city streets and fields that grow* *Thanksgiving unites a tapestry of faces* * In a shared celebration. * *Turkey, pie, and sides galore* * Tell stories of culture and tradition.* * While data whispers tales of who we are* *Our regions, trends, and shared embrace* *Yet beyond the numbers* * Lies the true heart of Thanksgiving-* *The laughter, love, and humble pride * *That illuminate the soul of this great land.* We will see you tomorrow, Day 2, with **Desserts** at the Thanksgiving table. Meanwhile, if you wish to play around with this dataset in Querri, you can download it below and try the following prompts: a. "Clean the data frame" b. "EDA" c. "Move the id column to the front of the data frame and convert it to an integer." d. "Make helpful suggestions on actions I can take on this dataset." Or give it a whirl with whatever thought sparks your curiosity! [Sign up for the Querri trial](https://app.querri.com) and download the data below. Note: You may not receive a confirmation after the download even though the file is downloaded. Check your downloads folder for filename TG Day 1 Data.csv file.