Research Paper ka Results Section: Kya Include Karein

Results section sabse aasan lagta hai. Aapke paas output hai, aap tables ko paste karte hain, aur jo hua use describe karte hain. Fir ek reviewer poochhta hai ki percentages ke denominators kyun nahi hain aur jo outcome aapne measure kiya tha wo kahin dikh kyun nahi raha.

Paanch failures, aur har ek ke neeche uska fix. Sabhi examples is guide ke liye models ke roop mein banaye gaye hain.

<CTA title="Table aur text ke beech numbers ko consistent rakhein" description="Apne results ko apne sources aur notes ke saath ek hi workspace mein draft karein, taaki text ke figures tables se match karein." buttonLabel="Try Jenni Free" link="https://app.jenni.ai/register" />

Ek research paper ke results section mein kya jata hai?

Bina kisi interpretation ke, jo aapko mila use ek logical order mein report karna. "Readmission 8.1 percentage points gir gaya" ek result hai. "Yeh dikhata hai ki pharmacist review kaam karta hai" ek discussion hai.

ICMJE Recommendations results ko logical sequence mein sabse important findings ke saath pehle dikhane ki maang karte hain, aur sirf statistical hypothesis testing par rely karne se mana karte hain, jo effect size aur precision ko sahi se nahi batata.

Failure 1. Percentage jiska koi denominator nahi hai

Kya likha jata hai: Intervention group ne controls ke muqable kam readmission rate dikhaya (22.4% vs 30.5%).

Kya likha hona chahiye: Intervention group mein 406 mein se 91 patients (22.4%) aur control group mein 406 mein se 124 patients (30.5%) ka readmission hua, jo ki 8.1 percentage points ka risk difference hai (95% CI 2.1 se 14.1, p = .008).

ICMJE numeric results ko na sirf derivatives jaise percentages ke roop mein mangta hai, balki un absolute numbers ke roop mein bhi mangta hai jisse ye derivatives calculate kiye gaye the. Ek reader bina yeh jaane 22.4% ko judge nahi kar sakta ki yeh 91 log hain ya sirf 9.

Failure 2. P value se hi saara kaam karwana

Ek p value yeh nahi batati ki effect kitna bada hai ya aapne ise kitne precision se estimate kiya hai. American Statistical Association ka 2016 statement cheh principles batata hai, jinme se teen yahan matter karte hain: p values is baat ki probability ko measure nahi karti ki hypothesis sach hai, conclusions is baat par depend nahi karne chahiye ki p value kisi threshold ko pass karti hai ya nahi, aur p value kisi effect ke size ko measure nahi karti.

Pehle estimate report karein, fir uska interval, aur fir p value. Isi order mein, kyunki interval mein wo sab hota hai jo reader ko chahiye.

Purdue OWL is table ke peeche ke do rules ko summarize karta hai: exact p values ko do ya teen decimal places tak report karein, aur leading zero ka use tabhi karein jab statistic 1 se zyada ho sake.

<ProTip title="🔢 Precision:" description="Kabhi bhi p = .000 report na karein. Software ise tab print karta hai jab value uski display precision se chhoti hoti hai, aur yeh aisi cheez batata hai jo sach nahi hai. Iski jagah likhein p is less than .001." />

Failure 3. Outcome jo chupchaap gayab ho gaya

Sirf unhi outcomes ko report karna jo significance tak pahuche hain, selective reporting hai, aur reviewers ab tezi se protocol ya registration check karte hain yeh dekhne ke liye ki kya chhoot gaya hai.

Har us outcome ka naam likhein jo aapne measure kiya hai, usi order mein jo aapne pehle se decide kiya tha, aur post hoc analyses ko post hoc label karein. CONSORT explicitly kisi bhi dusre analyses ke baare mein poochhta hai, jisme subgroup aur sensitivity analyses shamil hain, aur pre-specified ko post hoc se alag karta hai.

Failure 4. Wo paragraph jo Table 1 ko fir se sunata hai

ICMJE authors se kehta hai ki tables ke saare data ko text mein repeat na karein, aur sirf sabse important observations ko hi emphasize ya summarize karein. Ek aisa paragraph jo har ek baseline characteristic ko list karta hai, results section ko zaroorat se dugna lamba banane ka sabse aam tarika hai.

Iski jagah table ki taraf ishara karein. "Full baseline characteristics Table 1 mein hain" ek complete sentence hai aur yeh pura kaam kar deta hai.

Table, figure, ya text?

The European Association of Science Editors sabse saaf rule deta hai: tables tab jab exact numerical values matter karti hain, figures comparisons ya trends dikhane ke liye, aur dono mein kabhi bhi same data nahi hona chahiye.

• Table tab jab reader kisi specific value ko dekhna chahta ho, ya jab ek saath kai variables dikhane hon.

• Figure tab jab value se zyada shape matter karta ho.

• Text tab jab teen ya usse kam numbers hon. Ek do row ka table aamtaur par ek sentence hi hota hai jo chhupa hota hai.

Failure 5. Wo verb jo interpret karta hai

Interpreting: pharmacist review ka ek clear benefit demonstrate kar raha hai. Emergency attendance bhi favorable direction mein trend kar raha tha.

Reporting: Emergency department attendance groups ke beech alag nahi tha (18.2% vs 20.4%, risk difference 2.2 percentage points, 95% CI -2.6 se 7.0).

Pehle version mein do cheezein galat huin. "Demonstrating" ek discussion verb hai. Aur "trended in the favorable direction" us result ko describe karta hai jo hua hi nahi, jo ki results section mein spin ka sabse aam roop hai.

<ProTip title="📊 Reporting:" description="Apne results section ke har ek verb ko circle karein. Agar koi bhi verb demonstrates, proves, confirms ya suggests hai, toh us sentence ko discussion mein le jayein aur use us number se replace karein jise wo chhupa raha tha." />

Published results mein kitni baar galtiyan hoti hain

Behavior Research Methods mein ek study ne teen dashak ke psychology articles ka ek automated check chalaya. 30,717 articles jisme 258,105 p values thi, unme se aadhe papers jo significance testing use kar rahe the, unme kam se kam ek p value aisi thi jo uske apne test statistic aur degrees of freedom se match nahi kar rahi thi. Aath mein se ek mein grossly inconsistent value thi jo shayad conclusion ko affect kar sakti thi.

Yeh koi fraud nahi hai. Yeh transcription slips hain aur numbers jo ek jagah update hue par dusri jagah nahi, jiska matlab hai ki inhein kuch hi minutes mein pakda ja sakta hai.

<ProTip title="🧮 Math check karein:" description="Apne results ko apne output file ke saath line by line padhein, memory se nahi. In inconsistencies mein se lagbhag har ek copy error ke roop mein shuru hui thi." />

Jenni ke saath results draft karna

Prompt mein section ko describe karein. Study ka naam likhein aur kahein ki aapko intervals ke saath effect sizes chahiye, taaki reporting convention shuru se hi set ho jaye.

Citation style set karein. Statistical formatting conventions style ke saath hi chalti hain.

Structure generate karein. Standard IMRaD headings, taaki results apne khud ke section mein rahein aur interpretation ke paas kahin aur jaane ki jagah ho.

Fir outcome by outcome draft karein: pehle primary, fir secondary usi order mein jo aapne pre-specify kiya tha, aur subgroups ko saaf label karein. Statistics mein formatting slips ke liye Proofread chalayein aur un sentences ke liye Claim Confidence chalayein jahan reporting arguing mein badal gayi thi. Dono review pass mein hote hain.

Consistency pass

Pandraah minutes, aur yeh saare paanch failures ko pakad leta hai.

• Text mein har percentage ke saath uska denominator hota hai.

• Har effect ke paas ek interval hota hai, jisme null waale bhi shamil hain.

• Aapke methods mein bataya gaya har outcome aapke results mein kahin na kahin dikhta hai.

• Koi bhi paragraph us table ko fir se nahi batata jise aap pehle hi point kar chuke hain.

• Section ka koi bhi verb interpret nahi karta.

• Text ka har number us table se match karta hai jahan se wo aaya hai.

<ProTip title="🚦 Order:" description="Results ko usi order mein likhein jis order mein aapke methods ne outcomes ko introduce kiya tha. Agar reader ko primary outcome ke liye dhoondhna pade, toh ordering galat hai bhale hi baaki findings kitni bhi interesting kyun na hon." />

Numbers ko bolne dein

Ek accha results section jaanboojhkar aapke paper ka sabse kam persuasive hissa hota hai. Aap wahi report karte hain jo hua, usi order mein jisme aapne kaha tha ki aap ise measure karenge, itni precision ke saath ki ek dhyaan se padhne wala reader aapke discussion se disagree kar sake par fir bhi aapke data par trust kare. Yeh restraint hi discussion ko credible banata hai jab wo saamne aata hai.

<CTA title="Reviewer ke check karne se pehle reporting check karein" description="Draft par Proofread aur Claim Confidence chalayein taaki un sentences ko pakda ja sake jahan reporting arguing mein badal gayi thi." buttonLabel="Score My Draft" link="https://jenni.ai/cfp-review" />

Aage ke section ke liye, research paper ke discussion section ke hamare guide ko dekhein, aur us sabko summarize karne ke liye, conference abstract likhna dekhein. Agar paper abhi ban hi raha hai, toh apne sources ke against draft karne se numbers unki asli jagah ke paas rehte hain.