Read the sentence before counting its references
Copy the exact claim you want to check. Identify its action word: defines, classifies, measures, predicts, improves or treats. A page may cite a legitimate source while making a stronger claim than that source supports. Ten references at the bottom do not tell you which one supports the sentence.
MedlinePlus recommends checking who publishes and funds health information, its review process, its sources and its date. Those are useful quality questions. They do not remove the next step: reading what the linked source actually says about this specific claim.
Three common TCM citations need different conclusions
On a narrow screen, scroll this table sideways to read all columns.
| Source | A supported use | A leap it does not supply |
|---|---|---|
| CMRO traditional food guide | Attribute a named food's nature or flavor to that guide. | Its category independently proves a clinical treatment result. |
| A named CCMQ validation paper | Discuss the questionnaire and measurement properties actually studied. | A different online quiz inherits its accuracy, or a diet selected by the quiz works. |
| NCCIH TCM overview | Report its dated background on different practices and care cautions. | Every food, product or new claim has been individually reviewed. |
For example, the CMRO page describes mung bean as cold in nature. “This guide assigns that traditional category” is a traceable conclusion. “The category proves a treatment for my condition” is a separate clinical claim. Use the food-claim worksheet for the more detailed preparation, population, comparison and outcome checks.
Say how much of the source you actually read
Record whether you opened the full article, an abstract, one defined section or a secondary summary. A search snippet helps locate a source; it does not replace the source. If the needed methods or result are unavailable, write “not available in what I read.” Do not fill the gap from another article without naming that additional source.
Reading a review's reference list also does not mean that you read every paper on it. A review can be useful for the question it addresses; identify its scope and evidence dates. MedlinePlus notes that seemingly conflicting studies may have different designs and ask different questions.
Separate the publication date from the evidence date
A recently updated webpage may describe older research. NCCIH's TCM overview was last updated in April 2019. It is dated background here, not a newly completed food-by-food review. The 2013 and 2023 questionnaire papers concern their named versions and populations; a current link does not change those methods.
Keep the source's date, any stated review search date and your access date in different fields. Only record a date the source supplies. If newer evidence is relevant but has not been checked, leave that as an unresolved question rather than declaring the evidence current.
Finish with one bounded sentence
A useful record can say: “I read this section of this source. It supports this traditional classification, but not the treatment conclusion in the headline.” Another can say: “This paper studied a named short form; I have not seen testing of the different quiz I used.” These statements preserve both the information and its limit.
For a personal health decision, discuss the actual question with an appropriate professional. An accurate citation record helps the conversation; it does not select a treatment, verify an interaction or decide whether care can wait.
Match one statement to the source you actually read
Examples to work through
These fictional examples practice reading and keeping information accurate.
A citation supports the definition only
A fictional article correctly cites CMRO for a food nature and then adds a treatment promise. Keep the category as an attributed traditional description; the stronger promise still needs directly relevant evidence.
A questionnaire name is borrowed
A fourteen-question quiz cites the 2023 paper on SF-WQ9CCMQ-A. The studied form has 27 items. A citation to that paper does not validate the different questions, weights or online workflow.
Only the abstract was available
A reader has an abstract but not the full methods. They record that reading scope and the missing details. They do not describe the whole paper as reviewed or guess its preparation from the headline.
A new webpage contains old evidence
A recently edited page links to a dated overview. Record both dates separately. The edit date does not make the overview a current systematic review of every claim.
A private claim-and-citation note
Use these optional reminders for your own notes. The blank download contains all record fields. Keep a completed copy appropriately private; these controls do not submit entries.
Frequently asked questions
Does an official source prove every sentence on a website?
No. Check the exact statement and what the official source supports. Publisher credibility and claim matching are separate checks.
Must I read every linked reference?
Start with the reference supporting your specific question. Describe your actual reading scope; do not claim unread papers were reviewed.
What should I write when important details are missing?
Mark them as unavailable or not reported in what you read. Keep the uncertainty visible rather than completing a method from assumptions.
Where is the detailed food-study checklist?
Use the linked food-claim worksheet for preparation, participants, comparison, outcome, duration and research-context checks. This page explains citation scope across different TCM topics.
Sources and what they support
- MedlinePlus: Evaluating Health Information
Publisher, purpose, funding, review, dates, source tracing and discussion before personal health reliance.
- MedlinePlus: Understanding Medical Research
Study design, population, size, duration, harms and context when interpreting research reports.
- Hong Kong CMRO: Health Preservation by Food in Chinese Medicine
Attributed traditional nature/flavor categories, named food forms and a four-part constitution appendix. Its traditional effect wording is not treated here as independently demonstrated clinical efficacy.
- Wong et al. (2013): Validation of the Constitution in Chinese Medicine Questionnaire
The introduction, methods, discussion and conclusions were read for this repair: original 60-item scales, Hong Kong adaptation, overlap, score versus classification reproducibility and study limits. Other result tables and cited papers are not claimed independently reviewed here.
- Bai et al. (2023): Development and evaluation of a short-form CCMQ
The background, limitations and conclusions were read for this repair: the named 27-item SF-WQ9CCMQ-A, retained nine-scale structure and scope limits. Other results and cited papers are not claimed independently reviewed here.
- NCCIH: Traditional Chinese Medicine—What You Need To Know
Dated April 2019 background on different practices, product risks, coordinated care and avoiding replacement or delay of conventional care; not a current review of every food or product.
These sources support the stated explanations and reading limits. They do not establish a clinical review of this website, validate its fourteen-question sorter, or assess an individual product, symptom or treatment.