Dear ChatGPT. Help me Learn to Make a Pivot Table

In my conversation last week with Karen Kohn, we discussed the necessity of humans interacting with data to achieve the best analytical results. We were exploring statistics from EBSCO on the use of our e-books to understand the top subject areas within this platform.

I shared with Karen the ChatCPT-generated analysis of usage trends and she described her alternative tool of choice – the pivot table function within Excel. Pivots are an easy way of summarizing data – in this case over 5,000 records. But Karen also noted that many library staff are not experienced with Excel.

If the barrier to analysis is skill with Excel, is this something that ChatGPT could remedy? Yes, ChatGPT can help us to create learning materials. In this example, I asked ChatGPT to teach me how to create a pivot table for the same Excel file of e-book usage statistics (download here for Temple Library staff). 

Here was my prompt: 

Using this excel file, create a step by step tutorial for a novice data librarian on the creation of a pivot table to show the top subject areas by total usage. 

The result was pretty bare bones, but here are the first few steps (The full slide deck is available here):  

📊 Tutorial: Creating a Pivot Table for Top Subject Areas by Total Usage 
🧾 Step 1: Open Your Excel File 
Open the Excel file: 
eBookTitle_Jun042025_s8483726_20250301_20250531 (1).xlsx 
Identify the column labeled “BISAC” (subject area) and the column “Total Access” (usage). 
 
📍 Step 2: Select the Data Range 
Click on any cell in the dataset. 
Press Ctrl+A to select the entire dataset. 
Make sure there are column headers (e.g., “Title,” “BISAC,” “Total Access”). 
 
➕ Step 3: Insert a Pivot Table 
Go to the Insert tab on the top menu. 
Click on PivotTable
In the pop-up window: 
Ensure your full table range is selected. 
Choose to place the pivot table in a New Worksheet
Click OK
Note: All formatting provided by ChatCPT

For those who want to gain skills like this, we have many other tools (LinkedIn learning or other training are options). An advantage here is a more customized option, as ChatGPT’s tutorial utilized the file we specified.

What are the other differences? How should this “novice data librarian” select the best tool for a small assessment task like analyzing usage statistics? I asked ChatGPT:

Provide me with a list comparing the use of ChatGPT to analyze this data and the use of Excel and pivot tables to perform that analysis

ChatCPT came up with 14 examples. Here are a few:  

📊 Comparison: ChatGPT vs. Excel PivotTables for Data Analysis 

Feature / Task ChatGPT Excel PivotTables 
Ease of Use for Novices High – conversational guidance, no need to know formulas Moderate – intuitive GUI, but may require learning steps 
Setup Required Minimal – just upload file and ask questions Requires opening Excel, selecting data, inserting PivotTable 
Speed of Insight Fast – instant aggregation, summaries, and recommendations Fast – once PivotTable is set up, updates are automatic 
Customization of Output Very flexible – can sort, filter, summarize with plain text or charts Moderate – can pivot, filter, and chart but with less automation 
Text-Based Summaries Strong – can explain patterns and generate narratives Weak – user must interpret data visually or manually write summaries 
Visualization Options Can generate clean, customized charts Good – built-in chart tools, but with limited customization 
Error Checking / Guidance Yes – explains problems (e.g., wrong column types) Limited – errors are usually user-detected or implicit 
Note: Table formatting provided by ChatCPT. No editing.

It’s interesting that ChatGPT replicates my use of the word “novice” as the first difference. ChatGPT thinks very highly of itself. There’s nothing much it can’t do when compared side-by-side with Excel. It is indeed, easy to use, it’s fast, and no special software is required. ChatCPT is always willing to suggest patterns and recommendations.

Unlike Excel, where the user must interpret the data, ChatCPT is happy to provide an answer. And this is where some skepticism and critical thinking – the essential human element – are required.

Thanks again to Karen for prompting me to take my example one step further to share an additional use case for generative AI tools in support of assessment.

This entry was posted in process improvement, technology use and tagged , , , . Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.