# Reference Tracking Sheet Template By the end of this guide you will be able to build a working reference tracking sheet, populate it with references from a literature search, detect duplicate records automatically, record screening decisions for every source, and produce clean counts you can report in a thesis, manuscript, or systematic review. You need only a spreadsheet program such as Microsoft Excel, Google Sheets, or LibreOffice Calc. A reference manager such as Zotero or EndNote is helpful but optional, and this guide explains how a sheet and a manager work together. Everything here is written so you can follow it at a keyboard. Each step includes the exact text to type, the formula to paste, and the output you should expect to see. ## What a Reference Tracking Sheet Is A reference tracking sheet is a structured table, usually a spreadsheet, in which every reference you encounter during a literature search gets one row, and every fact you need about that reference gets one column. Those facts include the full citation, a unique identifier, where the reference came from, what you decided about it, and why. The sheet exists because literature searching involves more than running database queries. Current methodological guidance describes systematic searching as a combination of electronic database searching with supplementary search methods, and citation tracking is one of those supplementary methods [1]. Citation tracking means starting from one or more "seed references" and collecting the works they cite (backward tracking) and the works that cite them (forward tracking). A separate program of methodological research, including a scoping review and a Delphi consensus study, is working toward evidence-based recommendations on exactly how citation tracking should be used in health-related literature searching [1]. Until that guidance is final, the practical burden of organizing what you find falls on the searcher, and that is what the tracking sheet does. Published systematic reviews show the pattern you are imitating. A systematic review of the Shared Decision Making Questionnaire searched three databases (EMBASE, PsycINFO, and Medline) and combined that with reference tracking, citation tracking, and personal knowledge, with two independent reviewers screening titles, abstracts, and full texts [2]. A review of patient information-need measures ran a primary search in PubMed and Embase and then carried out a secondary search using reference and citation tracking of the included articles [3]. In both cases, the references found outside the primary database search had to be logged, deduplicated, screened, and decided on. A reference tracking sheet is the tool that makes this auditable. ## Why a Sheet Beats Memory and Loose PDFs Three practical reasons justify the ten minutes it takes to set up a sheet. **First, supplementary searching produces references that databases miss.** When a review team adds reference tracking and citation tracking to database searching, the new hits arrive scattered across sources: a bibliography in a review article, a citing paper found in a citation index, a guideline mentioned by a colleague. Without a sheet, these references live in browser tabs and download folders and are easily lost or accidentally counted twice. **Second, screening decisions need a written trail.** In the shared decision making review, data were extracted using a pilot tested sheet, and methodological quality was assessed with standardized tools [2]. That discipline starts earlier, at the reference level. When a co-author, examiner, or editor asks why a particular study was excluded, the answer should be one cell away from the citation, not buried in an email thread. **Third, counts matter for reporting.** Systematic reviews, including large evidence syntheses such as the Campbell Collaboration review of educational interventions for people with disabilities in low- and middle-income countries [4], report how many records were identified, screened, and included. You can only report those numbers if you counted them somewhere at the time, and a spreadsheet column of decisions gives you those counts with one formula. ## The Core Format: Columns in a Reference Tracking Sheet Template The reference tracking sheet template format below works for course papers, theses, and full systematic reviews. Add or remove columns to fit your project, but keep the first five in every version. | Column | Purpose | Example entry | |--|--|--| | ref_id | A unique code for each record so notes and files can link back | R014 | | authors_year | First author surname and publication year for quick scanning | Meier, 2021 | | title | Full article or document title | Citation tracking in health literature searching | | source | Journal, book, report, or database record | Systematic Reviews | | locator | DOI, PMID, or URL, so anyone can retrieve the item | Paste the DOI string here | | date_added | The date the record entered the sheet | 2026-11-11 | | found_via | How you found it: database name, backward reference tracking, forward citation tracking, hand search, or expert contact | Backward tracking from R003 | | screen_stage | Where the record sits: title, abstract, or full-text review | Full text | | decision | Include, exclude, or pending | Exclude | | reason | The exclusion reason or a note on inclusion | Animal study, not human participants | | key_finding | One sentence on what the source reports | Citation tracking added unique records beyond database searching | | notes | Quality comments, caveats, links to your PDF | PDF saved as R014.pdf | The `found_via` column deserves special attention. It is what distinguishes a reference tracking sheet from a simple bibliography. Because citation tracking collects cited and citing references from seed references [1], knowing how each record entered your project tells you whether your supplementary search is paying off, and it lets you report the contribution of each search method separately if your write-up requires it. ## Build Your Own Free Reference Tracking Sheet Template in Five Steps You do not need special software to make a free reference tracking sheet template. A blank spreadsheet becomes a research template in about ten minutes. The walkthrough below uses Google Sheets or Excel. The formulas are identical in both. ### Step 1: Choose your tool Open a blank spreadsheet. Name the file with the project name and a version date, for example `reftrack_handwashing_2026-11-11`. Keeping the date in the filename creates a simple version history you can fall back on. ### Step 2: Set up the header row In row 1, type the headers exactly as follows, one per cell, columns A through L: ``` ref_id | authors_year | title | source | locator | date_added | found_via | screen_stage | decision | reason | key_finding | notes ``` Freeze row 1 so headers stay visible while you scroll. In Google Sheets use View, then Freeze, then 1 row. In Excel use View, then Freeze Panes, then Freeze Top Row. ### Step 3: Add a duplicate-detection key Duplicates are the most common problem in reference logs, because the same paper arrives from a database, a bibliography, and a citation index with slightly different formatting. Create a hidden helper column, column M, titled `dup_key`. In cell M2, enter: ``` =LOWER(TRIM(B2))&"|"&LOWER(TRIM(C2))&"|"&D2 ``` This concatenates the author-year field, the title, and the source into one lowercase string. In cell N2, titled `dup_flag`, enter: ``` =IF(COUNTIF($M$2:M2,M2)>1,"DUPLICATE","") ``` Then copy both formulas down the column for as many rows as you expect to need, for example 2,000 rows. The `$` signs lock the start of the range so each row compares against every row above it. Expected output: when you paste in a reference that repeats an earlier author, title, and source, the `dup_flag` cell for the second copy shows the text `DUPLICATE`. Every unique record shows an empty cell. Sort or filter on column N to review flagged rows before deleting the later copies. ### Step 4: Add a screening counter block In cells P1 to P4, build a live tally so you always know where your search stands: ``` P1: Total records P2: =COUNTA(A2:A2000) P3: Excluded P4: =COUNTIF(I:I,"exclude") P5: Included P6: =COUNTIF(I:I,"include") P7: Pending P8: =COUNTIF(I:I,"pending") ``` Expected output: if your sheet holds 120 records with 74 excluded and 31 included, P2 reads 120, P4 reads 74, P5 reads 31, and P8 reads 15. These four numbers are exactly the figures you will need when you write your methods section or fill in a PRISMA-style flow description. ### Step 5: Populate the sheet and set a data-entry rule Enter each reference as one row. Set yourself one non-negotiable rule: no row is saved without a `found_via` entry. A row with no recorded origin is a row you cannot reconstruct later. To bring in records efficiently, export from the database you searched. In PubMed, for example, run your search, select the results, choose Send to, then choose File in CSV format, and open the downloaded file in your spreadsheet. Copy the columns that map to your headers, paste them in as new rows, and fill in `found_via` for the whole batch at once. ## A Reference Tracking Sheet Template Example with Real Rows Here is a reference tracking sheet template example showing how five rows look in practice. The titles are generic project examples rather than real papers, which is why the locator column shows placeholders you would replace with actual DOI or PMID strings. | ref_id | authors_year | title | found_via | screen_stage | decision | reason | |--|--|--|--|--|--|--| | R001 | Silva, 2022 | Effects of Drug X on kidney markers in rats | PubMed | Full text | Exclude | Animal model, review focuses on humans | | R002 | Okafor, 2023 | Drug X adherence in community clinics | PubMed | Full text | Include | Directly answers research question | | R003 | Larsen, 2020 | Systematic review of Drug X outcomes | Web of Science | Full text | Include | High-quality synthesis, seed reference | | R004 | Bianchi, 2021 | Childhood outcomes after Drug X exposure | Backward tracking from R003 | Title | Exclude | Outcome not measured | | R005 | Nunes, 2024 | Drug X dosing survey | Forward citation tracking from R003 | Abstract | Pending | Awaiting full-text retrieval | Notice three features of this example format. Each reference has an ID that appears nowhere else. The two rows found through citation tracking name the seed reference in `found_via`, consistent with the idea that citation tracking collects cited and citing references outward from seeds [1]. And the decisions map onto a staged pipeline: records start at title screening, advance to abstract and then full text, and receive a final decision with a reason. ## How the Sheet Fits a Systematic Review Workflow If you are conducting a formal systematic review, the sheet becomes part of a documented pipeline. The pattern across published reviews is consistent. The shared decision making review combined three database searches with reference tracking, citation tracking, and personal knowledge, then used two independent reviewers at every screening stage [2]. The information-needs review used a primary search in two databases and a secondary reference and citation tracking search, again with two independent reviewers screening titles, abstracts, and full texts [3]. In your sheet, support that workflow with three additions: 1. Add a `reviewer_1_decision` and `reviewer_2_decision` column pair, plus a `final_decision` column recording the resolved outcome. 2. Keep the `reason` vocabulary narrow. Pick a short fixed list, for example: wrong population, wrong outcome, wrong study design, not original research, no full text. Free-text reasons become impossible to count later. 3. Export a clean copy of the sheet when the search closes. That frozen file is the record of what you did, while the working file can keep evolving during writing. The counter block from Step 4 then gives you the numbers a review needs to report: records identified through database searching, records identified through reference and citation tracking, duplicates removed, records screened, records excluded with reasons, and studies included. ## Paper, PDF, and Hybrid Options ![Yes or no path that sends decision tracking to a spreadsheet and PDF storage and citation formatting to a reference manager.](/article-images/illustrations/reference-tracking-sheet-template-sheet-or-manager.webp) A spreadsheet is not the only way researchers track records, and it helps to understand the alternatives before choosing. **Paper registers.** Structured paper records still do real work in many settings. In an assessment of the VaxTrac electronic immunization registry in Sierra Leone, investigators compared the electronic system against three paper-based sources, including a register of children under two and a daily tally sheet, and treated the under-two register as the reference gold standard for evaluating data quality [5]. Two lessons transfer directly to reference management. First, a well-kept single register can serve as the authoritative record against which other copies are checked. Second, records that require capture at the moment of the event are the ones most prone to gaps, which is why your sheet should be updated when you find a reference, not at the end of the week from memory. Designate your spreadsheet as the master record and treat any sticky notes, printed PDF stacks, or folder names as secondary copies to be reconciled against it. **[Lab notebook](/knowledge/molecular-biology/lab-notebook) templates.** A lab notebook template pdf serves a different job from a reference tracking sheet. The notebook records what you did at the bench: dates, reagents, lot numbers, instrument settings, observations, and hand-drawn or printed results. The reference tracking sheet records what you read and what you decided about it. Many researchers keep both, and some institutions provide lab notebook template pdf files for printing and binding. If you prefer paper for the reading log as well, print the column headers from your spreadsheet as a table and rule off one block per reference. The tradeoff is that paper cannot run the duplicate formula from Step 3, so plan to transcribe paper entries into a spreadsheet before you analyze or report counts. **Reference managers.** Zotero, Mendeley, and EndNote store full records and PDFs and generate formatted bibliographies automatically. They are superior to a sheet for citation insertion and PDF storage. The sheet is superior for decision tracking, because managers are not built to record that record R004 was excluded at title screening for the reason "outcome not measured." The most efficient arrangement is to import everything into the manager for retrieval and storage, and maintain the sheet as the decision log. Export a bibliography from the manager periodically and check that every entry appears in the sheet with a matching `locator` value. ## Checking That Your Sheet Is Correct Before you trust your counts, run four checks. **Duplicate sweep.** Filter column N for `DUPLICATE` and confirm each flag is a genuine repeat. Two different papers can share a first author and year, which is why the key includes title and source. If a flag is wrong, refine the key by adding more of the title. **Count reconciliation.** Confirm that included plus excluded plus pending equals the total in P2. If it does not, some rows have a blank or misspelled decision. The formula `=COUNTIF(I:I,"")` reveals empty decision cells, and a filter on column I will surface typos such as "Include" with a capital letter, which the counter does not catch. **Locator spot check.** Pick five random rows and confirm the DOI, PMID, or URL in `locator` opens the correct document. Broken locators are easy to fix now and painful during writing. **Coverage check.** Sort by `found_via` and confirm each search method you planned to use actually contributed rows. If backward reference tracking from your seed articles produced nothing, either the seeds were poorly chosen or the column was left blank, and you need to know which before you write your methods. ## Common Mistakes and Limitations - **Recording the reference but not the decision.** A sheet full of citations with an empty `decision` column is a bibliography, not a tracking tool. The decision and reason columns are the reason the sheet exists. - **Skipping the unique ID.** Without `ref_id`, you cannot name PDF files, cite rows in emails to co-authors, or link the sheet to your extraction forms. Assign IDs the moment a row is created. - **Free-text exclusion reasons.** "Not relevant," "wrong sort of patients," and "did not fit" all mean different things to different readers six months later. Use a fixed reason list. - **Deduplicating by eye.** Manual deduplication fails on papers retrieved twice with different title capitalization or journal abbreviations. The COUNTIF key from Step 3 catches these mechanically. - **Letting the sheet and the reference manager drift apart.** If you delete a row in the sheet but the PDF stays in your manager, or the reverse, your final bibliography and your reported counts will disagree. Reconcile at defined points, such as after each search update. - **Treating the sheet as the whole method.** A tracking sheet organizes what you find. It does not design your search, validate your inclusion criteria, or substitute for the structured, multi-stage processes that systematic reviews follow, including dual independent screening [2] and combined database plus supplementary searching [1]. For a thesis or formal review, follow your institution's or journal's reporting requirements, and ask your supervisor or librarian to review your search plan before you finalize it. ## Frequently Asked Questions ### What is a reference tracking sheet template? It is a pre-structured spreadsheet that gives every reference found during a literature search one row, with fixed columns for the citation, its origin, the screening stage, and your include or exclude decision. You fill it in as you search so no source is lost or counted twice. ### What columns should the format include? At minimum, a unique reference ID, author and year, title, source, a DOI or PMID locator, how you found the reference, the screening stage, the decision, and the reason for the decision. Add columns for key findings and notes when you are working toward a formal review. ### Where can I get a free reference tracking sheet template? You can build one free in any spreadsheet program by following the five steps in this guide, which take about ten minutes. Because projects differ in their columns and reason lists, a template you build yourself from this format usually fits better than a downloaded one, and there is nothing to install. ### How is reference tracking different from citation tracking? Reference tracking generally means working backward through the reference lists of known papers to find earlier sources, while citation tracking also includes working forward through indexes to find papers that cite your seed references. Methodological work on citation tracking treats it as a supplementary search method that collects cited and citing references from seed references, and formal recommendations on its use are still being developed through a scoping review and Delphi study. ### Can I use a spreadsheet instead of a reference manager? For the decision-tracking job, yes, and a spreadsheet does things a manager cannot, such as recording exclusion reasons and generating live screening counts. For storing PDFs and inserting formatted citations while you write, a reference manager is the better tool, so most researchers use both together. ### How does the sheet fit into a systematic review? It serves as the auditable log behind your reported numbers. Published reviews combine database searching with reference and citation tracking and use two independent reviewers at each screening stage, and your sheet is where those parallel decisions, exclusions, and reasons live before they are summarized in your results. ### Is this the same "tracking" used in cell biology and imaging papers? No. In the life sciences literature, the word tracking also refers to following objects or features over time, such as computationally tracking epithelial cells across [fluorescence microscopy](/knowledge/diagnostics/imaging/fluorescence-microscopy-principles-applications-and-image-acquisition) frames, cardiac strain tracking on magnetic resonance images, and cell tracking in time-lapse light sheet imaging of embryos. Those methods involve software analysis of image data, not literature management, so do not mix them up when searching for templates. ### Do I need a reference tracking sheet for a class paper or thesis? For a short class paper with a dozen sources, careful use of a reference manager is usually enough. For a thesis chapter, a scoping review, or any project where you must explain how sources were found and selected, the sheet pays for itself immediately, because it turns your search process into a record you can defend. ## Further Reading - [4D light sheet imaging, computational reconstruction, and cell tracking in mouse embryos.](https://pubmed.ncbi.nlm.nih.gov/39754721/) ## Sources 1. [Using citation tracking for systematic literature searching - study protocol for a scoping review of methodological studies and a Delphi study.](https://pubmed.ncbi.nlm.nih.gov/34631036/) 2. [Use of the 9-item Shared Decision Making Questionnaire (SDM-Q-9 and SDM-Q-Doc) in intervention studies-A systematic review.](https://pubmed.ncbi.nlm.nih.gov/28358864/) 3. [Assessment of patient information needs: A systematic review of measures.](https://pubmed.ncbi.nlm.nih.gov/30703103/) 4. [Effectiveness of interventions for improving educational outcomes for people with disabilities in low- and middle-income countries: A systematic review.](https://pubmed.ncbi.nlm.nih.gov/39917627/) 5. [Assessment of VaxTrac electronic immunization registry in an urban district in Sierra Leone: Implications for data quality, defaulter tracking, and policy.](https://pubmed.ncbi.nlm.nih.gov/32753291/)