I'm conducting a systematic literature review on the effectiveness of mindfulness-based interventions for anxiety in college students. Help me: 1) Develop a comprehensive search strategy for PubMed and PsycINFO, 2) Create inclusion/exclusion criteria, 3) Design a data extraction form, 4) Suggest quality assessment tools (PRISMA, Cochrane Risk of Bias)

Prompt
I'm conducting a systematic literature review on the effectiveness of mindfulness-based interventions for anxiety in college students. Help me: 1) Develop a comprehensive search strategy for PubMed and PsycINFO, 2) Create inclusion/exclusion criteria, 3) Design a data extraction form, 4) Suggest quality assessment tools (PRISMA, Cochrane Risk of Bias)
DeepSeekDeepSeek
I'm conducting a systematic literature review on the effectiveness of mindfulness-based interventions for anxiety in college students. Help me: 1) Develop a comprehensive search strategy for PubMed and PsycINFO, 2) Create inclusion/exclusion criteria, 3) Design a data extraction form, 4) Suggest quality assessment tools (PRISMA, Cochrane Risk of Bias)

Okay, so I need to help someone with their systematic literature review on mindfulness-based interventions for anxiety in college students. Let me start by breaking down each of their four requests.

First, the search strategy for PubMed and PsycINFO. I remember that systematic reviews require comprehensive searches, so I should think about the key concepts: mindfulness-based interventions, anxiety, and college students. For each database, I'll need to use controlled vocabulary like MeSH terms for PubMed and maybe the Thesaurus terms for PsycINFO. Also, including synonyms and related terms is important. Maybe use Boolean operators to combine these terms. For example, "mindfulness" could have terms like MBCT, MBSR, or "mindfulness-based stress reduction". Anxiety terms might include "anxiety disorders", "generalized anxiety disorder", etc. College students could be "students", "university", "undergraduate", etc. Also, setting limits like publication date (last 10 years?), language (English?), and age group (young adults?).

Next, inclusion and exclusion criteria. They need to define what studies they include. Population: college students with anxiety. Maybe include both diagnosed and non-diagnosed? Intervention: mindfulness-based interventions, like MBSR, MBCT, or other structured programs. Comparison could be active controls, waitlist, or no treatment. Outcomes: anxiety measures, validated scales. Study designs: RCTs, maybe quasi-experimental? Exclude non-peer-reviewed, non-English, other populations (like high school or adults not in college), interventions that are not mindfulness-based (like other therapies), studies without control groups, or those not measuring anxiety as a primary outcome.

Data extraction form. This should capture study details, population characteristics, intervention details, outcomes, results. So fields like author, year, country, study design, sample size, age, gender, inclusion/exclusion criteria, intervention type, duration, control group type, outcome measures, time points, main findings, limitations, funding source. Maybe also adverse events if any.

Quality assessment tools. PRISMA is for reporting guidelines, but they might need to mention they followed PRISMA. For risk of bias, Cochrane RoB 2.0 for RCTs. If there are non-randomized studies, maybe ROBINS-I. Also, maybe check for other quality aspects like CONSORT for reporting of trials. Maybe mention using tools like GRADE for assessing evidence quality, but Cochrane RoB is more about study methodology.

Wait, the user mentioned PRISMA and Cochrane Risk of Bias. So PRISMA is a guideline for conducting and reporting systematic reviews, so they should follow the PRISMA checklist. The flow diagram is part of that. Cochrane Risk of Bias tool is for assessing the methodological quality of included studies, specifically RCTs. So for each study, they'll assess domains like randomization, allocation concealment, blinding, incomplete data, selective reporting, other biases.

Let me make sure I cover all parts. For the search strategy, example terms for each database. For inclusion/exclusion, clear criteria. Data extraction form with all necessary fields. Quality assessment using Cochrane RoB 2.0 and mention PRISMA for reporting.

I should also think about possible filters. For example, in PubMed, using the Clinical Queries filter for therapy or systematic reviews. But since they're looking for original studies, maybe not. Also, check if they want only RCTs or other designs. The inclusion criteria should specify that. Maybe include RCTs and quasi-experimental if they allow it.

Another thing: when developing the search strategy, using truncation and wildcards for terms. Like "student*" to capture student, students, etc. Also, combining terms with OR within each concept and then AND between concepts. For example, (mindfulness terms) AND (anxiety terms) AND (college student terms). Testing the search strategy for sensitivity and specificity might be needed, but maybe that's beyond the scope here.

Inclusion criteria should have time frame? Maybe last 10 years unless they want all available. Also, language restrictions. Exclusion might include non-English studies, but that's a limitation. They need to justify that.

Data extraction form should be pilot-tested. Maybe include a field for notes or comments. Also, key outcomes like mean differences, effect sizes, p-values, confidence intervals. Maybe whether the study found a significant reduction in anxiety.

Quality assessment: Cochrane RoB 2.0 has specific questions for each domain. Each study is rated as low risk, some concerns, or high risk. They can summarize this in a table.

I think that's the main points. Now, structure each part step by step, making sure each of the four sections is addressed clearly with examples where possible.

1. Comprehensive Search Strategy


Databases: PubMed and PsycINFO
Key Concepts: Mindfulness-based interventions (MBIs), anxiety, college students.

PubMed Search Strategy


``
("mindfulness"[MeSH] OR "mindfulness-based stress reduction"[MeSH] OR "mindfulness-based cognitive therapy"[MeSH] OR "mindfulness intervention*"[tiab] OR MBSR[tiab] OR MBCT[tiab])
AND
("anxiety"[MeSH] OR "anxiety disorders"[MeSH] OR "generalized anxiety disorder"[MeSH] OR "social anxiety"[tiab] OR "anxiety symptom*"[tiab])
AND
("students"[MeSH] OR "college students"[tiab] OR "university students"[tiab] OR "undergraduate"[tiab] OR "young adult"[tiab])
Filters:
  • Publication date: Last 10 years (2013–2023)

  • Language: English

  • Age: 18–25 years

`

PsycINFO Search Strategy


`
(DE "Mindfulness" OR DE "Mindfulness Based Interventions" OR TIAB "mindfulness-based stress reduction" OR TIAB "mindfulness-based cognitive therapy" OR TIAB "MBSR" OR TIAB "MBCT")
AND
(DE "Anxiety" OR DE "Generalized Anxiety Disorder" OR TIAB "anxiety disorder" OR TIAB "social anxiety" OR TIAB "anxiety symptom")
AND
(DE "College Students" OR DE "Undergraduate Students" OR TIAB "university student" OR TIAB "young adult")
Filters:
  • Peer-reviewed journals

  • Publication date: 2013–2023

  • Language: English

``

Notes:

  • Use truncation (*) and proximity operators (e.g., TIAB for title/abstract).

  • Adapt terms to database-specific thesauri (MeSH in PubMed, Thesaurus in PsycINFO).

  • Export results to citation managers (e.g., EndNote, Zotero) for deduplication.



2. Inclusion/Exclusion Criteria


| Category | Inclusion Criteria | Exclusion Criteria |
|---------------------|---------------------------------------------------------------------------------------|---------------------------------------------------------------------------------------|
| Population | College/university students (undergraduate/graduate) with anxiety symptoms or diagnosis | Non-college populations (e.g., high school, general adults) |
| Intervention | Structured MBIs (e.g., MBSR, MBCT, mindfulness courses ≥4 sessions) | Non-structured mindfulness (e.g., self-help apps, yoga without mindfulness component) |
| Comparison | Active control (e.g., CBT), waitlist, or treatment-as-usual | No control group |
| Outcomes | Validated anxiety scales (e.g., GAD-7, STAI, DASS-21) | Studies not reporting anxiety outcomes |
| Study Design | RCTs, quasi-experimental designs | Case studies, qualitative studies, reviews |
| Language/Date | English, published 2013–2023 | Non-English, pre-2013 |


3. Data Extraction Form


Microsoft Excel/Google Sheets Template:
| Field | Description |
|---------------------------|---------------------------------------------------------------------------------|
| Study Details | Author, year, country, funding source, conflicts of interest |
| Population | Sample size, age (mean ± SD), % female, inclusion/exclusion criteria |
| Intervention | Type of MBI (e.g., MBSR), duration (weeks), frequency, delivery mode (group/online) |
| Comparison | Control group type (e.g., waitlist, active treatment) |
| Outcomes | Anxiety measures used, time points (pre/post/follow-up), effect size (Cohen’s d), significance (p-value) |
| Results | Key findings (e.g., "MBI reduced anxiety scores by 20% vs. control") |
| Limitations | Study limitations (e.g., small sample, attrition bias) |

Notes: Pilot-test the form with 2–3 studies to refine fields.


4. Quality Assessment Tools


  • PRISMA 2020 Checklist: Use for transparent reporting (e.g., flow diagram, abstract structuring).

  • Cochrane Risk of Bias 2.0 (RoB 2): Assess RCTs across 5 domains:

  • Randomization process

  • Deviations from intended interventions

  • Missing outcome data

  • Outcome measurement

  • Selective reporting

{DateTime.UtcNow:HH:mm}
GrokxAI
I'm conducting a systematic literature review on the effectiveness of mindfulness-based interventions for anxiety in college students. Help me: 1) Develop a comprehensive search strategy for PubMed and PsycINFO, 2) Create inclusion/exclusion criteria, 3) Design a data extraction form, 4) Suggest quality assessment tools (PRISMA, Cochrane Risk of Bias)

Below is a structured response to your query for conducting a systematic literature review (SLR) on the effectiveness of mindfulness-based interventions (MBIs) for anxiety in college students. I'll address each of your four requests step by step, drawing on standard SLR methodologies (e.g., from the PRISMA guidelines and Cochrane Handbook). This is tailored to your topic, but you may need to refine it based on your specific research question (e.g., "What is the effectiveness of MBIs in reducing anxiety symptoms among college students?").

1. Comprehensive Search Strategy for PubMed and PsycINFO


A robust search strategy should use a combination of keywords, MeSH terms (for PubMed), and subject headings (for PsycINFO), along with Boolean operators (AND, OR, NOT). Focus on three key concepts: Mindfulness-based interventions, Anxiety, and College students. Aim to capture variations in terminology (e.g., "university students," "undergraduates").
  • General Tips:

  • Run searches in both databases for comprehensiveness (PubMed for biomedical literature; PsycINFO for psychological/behavioral studies).

  • Use truncation (e.g., anxi* for anxiety/anxious) and synonyms.

  • Limit to English-language, peer-reviewed articles from the last 20–25 years (e.g., 2000–present) unless your scope differs.

  • Export results to a reference manager like EndNote or Zotero for deduplication.

  • Pilot the search and iterate based on results (target 500–2000 initial hits).


PubMed Search Strategy


PubMed uses MeSH (Medical Subject Headings). Here's a sample string (copy-paste ready; adapt date filters via the interface):

("Mindfulness"[Mesh] OR "Meditation"[Mesh] OR mindfulness* OR "mindfulness-based" OR MBSR OR MBCT OR "mindful* intervention*" OR "mindful* therapy" OR "mindful* practice*") 
AND
("Anxiety Disorders"[Mesh] OR "Anxiety"[Mesh] OR anxi* OR "panic disorder*" OR "generalized anxiety" OR GAD)
AND
("Students"[Mesh] OR "Universities"[Mesh] OR college* OR universi* OR undergrad* OR postgrad* OR "higher education" OR "young adult*" OR "emerging adult*")

  • Expected Yield: ~200–500 results (depending on date filters).

  • Additional Filters: Humans (not animals); RCT, quasi-experimental, or observational studies; 2000–2024.

  • Sensitivity Check: Add terms like "yoga" or "tai chi" if including related mindfulness practices.


PsycINFO Search Strategy


PsycINFO uses its own subject headings. Sample string (via EBSCO interface):

(TI (mindfulness* OR "mindfulness-based" OR MBSR OR MBCT OR "mindful* intervention*" OR "mindful* therapy" OR "mindful* practice*") OR AB (mindfulness* OR "mindfulness-based" OR MBSR OR MBCT OR "mindful* intervention*" OR "mindful* therapy" OR "mindful* practice*")) 
AND
(TI (anxi* OR "anxiety disorder*" OR "panic disorder*" OR "generalized anxiety" OR GAD) OR AB (anxi* OR "anxiety disorder*" OR "panic disorder*" OR "generalized anxiety" OR GAD) OR SU ("Anxiety" OR "Anxiety Disorders"))
AND
(TI (college* OR universi* OR undergrad* OR postgrad* OR "higher education" OR "young adult*" OR "emerging adult*") OR AB (college* OR universi* OR undergrad* OR postgrad* OR "higher education" OR "young adult*" OR "emerging adult*") OR SU ("College Students" OR "Universities"))

  • Expected Yield: ~300–600 results.

  • Additional Filters: Peer-reviewed; methodology limited to treatment outcome, clinical trial, or quantitative study; 2000–2024.

  • Expansion: Search gray literature (e.g., Google Scholar, theses via ProQuest) and hand-search key journals like Mindfulness or Journal of American College Health. Register your protocol on PROSPERO for transparency.


2. Inclusion and Exclusion Criteria


These criteria should be predefined in your protocol to minimize bias (PICO framework: Population, Intervention, Comparison, Outcome). Apply them during title/abstract screening (e.g., via Rayyan or Covidence) and full-text review.

Inclusion Criteria


  • Population: College/university students (undergraduate or graduate, aged 18–25 typically); must include a subgroup analysis for students if a mixed sample.

  • Intervention: Mindfulness-based interventions (e.g., MBSR, MBCT, mindfulness meditation programs); delivered in any format (group, individual, online) and duration (≥4 weeks).

  • Comparison: Any (e.g., waitlist control, active control like relaxation training, or no intervention); include uncontrolled studies if RCTs are scarce.

  • Outcomes: Primary: Anxiety reduction (measured by validated scales like GAD-7, STAI, or Beck Anxiety Inventory). Secondary: Related outcomes like stress, depression, or well-being.

  • Study Design: Randomized controlled trials (RCTs), quasi-experimental, cohort, or pre-post studies.

  • Other: Published in English; full-text available; from 2000 onward to capture modern MBIs.


Exclusion Criteria


  • Population: Non-student samples (e.g., general adults, K-12 students, clinical populations without student focus).

  • Intervention: Non-mindfulness interventions (e.g., pure CBT without mindfulness, unrelated therapies like acupuncture); very brief (<4 weeks) or non-empirical (e.g., protocols without outcomes).

  • Outcomes: No anxiety-related measures; only qualitative or theoretical papers.

  • Study Design: Case studies, reviews, commentaries, or non-empirical work (e.g., opinion pieces).

  • Other: Non-English; duplicates; retracted articles.


Aim for two independent reviewers to screen; resolve disagreements via discussion or a third reviewer. Expect 10–20% of initial hits to reach full-text stage.

3. Data Extraction Form


Design a standardized form (e.g., in Excel, Google Sheets, or Covidence) to extract key data from included studies. This ensures consistency and facilitates synthesis (narrative or meta-analysis). Extract data in duplicate for reliability.

Key Sections and Fields


  • Study Identification:

  • Authors, year, title, journal, DOI/PMID.

  • Country/setting (e.g., US university).

  • Study Characteristics:

  • Design (e.g., RCT, pre-post).

  • Sample size (total, completers, per group).

  • Population details: Age (mean/SD), gender (% female), ethnicity, baseline anxiety level.

  • Recruitment method (e.g., campus flyers).

  • Intervention Details:

  • Type (e.g., MBSR, app-based mindfulness).

  • Duration (weeks/sessions), frequency (e.g., 8 weeks, 2x/week).

  • Delivery (in-person, online), facilitator (trained instructor?).

  • Fidelity checks (e.g., adherence monitoring).

  • Comparison Group:

  • Type (e.g., waitlist, TAU).

  • Details (duration, content).

  • Outcomes:

  • Primary: Anxiety measure (tool, pre/post/follow-up scores, effect size e.g., Cohen's d).

  • Secondary: Other measures, adverse events.

  • Follow-up time (e.g., 3 months post-intervention).

  • Statistical analysis (e.g., ITT vs. per-protocol, p-values, confidence intervals).

  • Quality/Risk Notes:

  • Funding source, conflicts of interest.

  • Attrition rate (% dropout).


Pilot the form on 2–3 studies to refine it. For meta-analysis, prioritize extractable effect sizes.

4. Quality Assessment Tools


Assess study quality/risk of bias early in the process to inform synthesis (e.g., weight higher-quality studies more). Use tools appropriate for your designs; report in a summary table.
  • PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses):

  • Not a quality tool per se, but use the PRISMA 2020 checklist and flow diagram for reporting your SLR. It ensures transparency (e.g., track search yield, screening decisions, reasons for exclusions). Download the checklist from prisma-statement.org. Essential for publication.

  • Cochrane Risk of Bias (RoB 2) Tool:

  • Ideal for RCTs (most relevant for effectiveness studies). Assesses five domains: randomization process, deviations from intervention, missing data, outcome measurement, and selection of reported results. Rate each as low/high/some concerns; generate an overall risk score.

  • For non-RCTs (e.g., quasi-experimental), use the parallel tool RoB-NObS or adapt.

  • How to Apply: Two reviewers score independently; use Review Manager (RevMan) software. Example: High risk if allocation concealment is unclear.

  • Additional Suggestions:

  • ROBINS-I (Risk Of Bias In Non-randomized Studies - of Interventions): For quasi-experimental or cohort studies; compares to RCTs across 7 domains (e.g., confounding, bias in classification of interventions).

  • MMAT (Mixed Methods Appraisal Tool): If including mixed-methods studies (quick for diverse designs).

  • GRADE (Grading of Recommendations Assessment, Development and Evaluation): For overall evidence quality (high/moderate/low/very low) post-synthesis, focusing on inconsistency, imprecision, etc.

  • Threshold: Exclude studies with high overall risk

{DateTime.UtcNow:HH:mm}

Analysis

This comparison demonstrates the different approaches each AI model takes when responding to the same prompt. Here are the key differences observed:

Response Characteristics

DeepSeek: Provides a direct response with 112 sentences.

Grok: Provides a direct response with 144 sentences.

Key Takeaways

  • Each model brings unique strengths to this type of query
  • Response styles vary significantly between models
  • Consider your specific use case when choosing between these models

Try This Comparison Yourself

Want to test these models with your own prompts? Visit SNEOS.com to compare AI responses side-by-side in real-time.


This comparison was generated using the SNEOS AI Comparison ToolPublished: October 15, 2025 | Models: DeepSeek, Grok