Story mix, sources and sensational devices in DD News, Aaj Tak and NDTV India evening bulletins
Mapped to your syllabusUGC CBCS template CBCS · BJ-651 Media Project (documentary / short film / campaign + report + viva) · Semester 6Indian television news is dominated by two very different institutional logics. DD News, the public broadcaster's news channel under Prasar Bharati, carries a statutory public-service mandate and no advertising-driven ratings race. Aaj Tak and NDTV India are private channels competing in a market where audiences are measured by BARC India and prime-time windows must hold viewers. This MJMC dissertation asks a concrete question: when all three put out a prime-time Hindi news bulletin on the same evening, do they actually cover the day differently?
The study is a quantitative content analysis of the longest continuous scheduled news bulletin (not a debate or chat show) broadcast between 8 and 10 pm by each of the three channels on 14 sampled days. Sampling follows the constructed-week method — two constructed weeks drawn from a 12-week study window — giving roughly 42 bulletins and an estimated 400-plus news stories coded at story level.
Each story is coded with a 14-variable codebook: beat, geography, story type (anchor read, reporter package, live cross), number and type of sources quoted (official, party, expert, citizen, none), story length in seconds, and four sensational form devices (breaking-news graphics, alert ticker, dramatic music, dramatic replay), operationalised from Grabe, Zhou and Barnett's classic explication of sensationalism. Reliability is established with a second coder and Cohen's kappa. Chi-square, ANOVA and descriptive statistics then answer five research questions about beat mix, source diversity, airtime, sensational packaging and international coverage.
Syllabus alignment
Criticism of Indian television news concentrates on the prime-time debate — the shouting panel, the breaking-banner war — and treats the evening bulletin as neutral ground. Yet the bulletin is where a channel's routine news decisions live: what makes the running order, who gets quoted, how long a story runs, and whether packaging devices manufacture urgency. No reliable, publicly available comparison tells students and viewers whether DD News's public-service mandate produces a measurably different bulletin from Aaj Tak's or NDTV India's market-driven output on ordinary days — not election nights, not crisis weeks. Without such evidence, classroom claims about 'the public broadcaster covers differently' remain opinion, and media-literacy teaching lacks a current, method-clean data point. This dissertation fills that gap with a constructed-week sample, a published codebook, a tested reliability procedure and standard statistical tests, producing findings that are replicable by the next student and defensible before an external examiner.
This dissertation compares what the Indian public broadcaster and two private channels treat as news at the same hour of the same evenings. Using two constructed weeks sampled from a 12-week window, all stories in the longest scheduled prime-time Hindi bulletin (8–10 pm start, no debate shows) of DD News, Aaj Tak and NDTV India are coded for beat, geography, format, sources, duration and sensational form devices. Analysis uses frequencies, chi-square tests of independence, one-way ANOVA on story length, and Cohen's kappa for inter-coder reliability. The findings describe how institutional mandate and market pressure translate into measurable differences in the everyday bulletin.
Gatekeeping theory (Shoemaker and Vos) holds that every news organisation filters events through routines, ownership and audience pressures; agenda-setting research adds that whatever survives those gates frames what the public thinks about. In India these gates differ sharply by institution: Prasar Bharati broadcasters work under a public-service mandate with public funding, while private channels operate in a ratings ecosystem run by BARC India. Whether that difference still shows in the plain evening bulletin — as opposed to the heavily studied 9 pm debate — is an empirical question this study answers with numbers rather than impressions.
Indian content research has examined election front pages, newspaper framing and prime-time debates in depth, and Grabe, Zhou and Barnett (2001) supply a tested explication of sensationalism as content plus form. But everyday bulletins — the spine of a news channel's day — are rarely compared across the public–private divide with a reliability-tested codebook. Student studies often sample convenient single weeks and skip reliability checks entirely; this design fixes both weaknesses.
Two constructed weeks are drawn from the published schedules of the three channels during a 12-week study window. For each sampled day, the longest continuous scheduled news bulletin starting between 8 and 10 pm is recorded from each channel's official archive (bulletins of delayed upload are substituted per a pre-fixed rule). Every story in every bulletin is coded on 14 variables. Five research questions follow: RQ1 — do the channels differ in beat mix? RQ2 — how many sources does an average story carry, and of what type? RQ3 — how is bulletin airtime distributed across story lengths? RQ4 — how prevalent are the four sensational form devices? RQ5 — what share of coverage is international?
All three channels upload or archive their bulletins publicly, so the sample is buildable from a laptop without paid databases. Coding roughly 420 stories at 15–20 minutes each fits a 16-week academic schedule with a two-coder reliability phase. SPSS is available in the department computer lab, and every test used is taught in the BJ-403 Communication Research course. No human subjects are involved; recordings are used for academic analysis only, consistent with fair dealing for research under Section 52 of the Indian Copyright Act.
In scope
Out of scope
Design. Quantitative, comparative content analysis of television news, following Riffe, Lacy, Watson and Fico's constructed-week logic: content varies systematically by day of week, so two constructed weeks (one day of each weekday drawn randomly across the 12-week window, repeated for a second construct) represent the period better than any two consecutive weeks.
flowchart TD
A["Define 12-week study window and channel list"] --> B["Construct two sample weeks from published schedules"]
B --> C["Record flagged bulletins of DD News, Aaj Tak, NDTV India from official archives"]
C --> D["Pilot-code 10 percent and refine the codebook"]
D --> E["Full story-level coding, 14 variables"]
E --> F["Reliability: second coder re-codes 20 percent, Cohen's kappa"]
F --> G["Analysis: frequencies, chi-square, ANOVA / Kruskal-Wallis"]
G --> H["Report: tables, charts, discussion, replication log"]Sampling frame. Each channel's published weekly schedule fixes which programme is 'the longest continuous scheduled news bulletin starting 8–10 pm'; the rule is written before sampling begins and frozen after the pilot.
Instrument. A 14-variable story-level codebook: identification (date, channel, bulletin, story order, length in seconds); content (beat: politics, crime, business, international, sports, entertainment, health/education/environment, other; geography: national, state, international, local; format: anchor read, reporter package, live cross, desk interaction); sourcing (count and primary type: government/official, party spokesperson, expert/analyst, citizen/eyewitness, celebrity, none); and four form devices (breaking-news graphic, alert ticker, dramatic music or sting, dramatic slow-motion replay).
Reliability and analysis. A second coder (a classmate trained on the codebook) independently re-codes a random 20 percent subsample; kappa is computed per variable, and any variable below 0.70 triggers a codebook revision and recode before full analysis. Analysis runs in SPSS: frequencies per channel, chi-square for channel × beat and channel × device, one-way ANOVA (or Kruskal–Wallis if variances misbehave) on story length, and mean sources per story with confidence intervals.
| Phase | Weeks | Output |
|---|---|---|
| Design, schedules, sampling frame | 1–2 | Sampling plan, frozen bulletin rule |
| Pilot coding and codebook revision | 3–4 | Version 2 codebook, pilot kappa table |
| Recording and data entry | 4–8 | 42-bulletin archive log, coded sheet |
| Reliability phase | 8–9 | Final kappa table |
| Analysis | 10–12 | SPSS outputs, five RQ tables and charts |
| Writing and submission | 12–16 | Dissertation draft, similarity report, viva PPT |
As a research project, the 'architecture' is the study design — how one evening of television becomes a defensible dataset — and the structure of the codebook around the unit of analysis, the story.
flowchart LR
SCH["Published channel schedules"] --> RULE["Frozen rule: longest scheduled bulletin, 8-10 pm start, no debates"]
RULE --> ARCH["Official archives: DD News, Aaj Tak, NDTV India"]
ARCH --> CODE["Story-level coding sheet, 14 variables"]
CODE --> REL["Reliability: 20 percent re-coded, Cohen's kappa >= 0.70"]
REL --> SPSS["SPSS: chi-square, ANOVA, descriptives"]
SPSS --> RPT["Five research questions answered in tables and charts"]flowchart TD
S["Story — unit of analysis"] --> ID["Identification: date, channel, order, length in seconds"]
S --> CT["Content: beat, geography, format"]
S --> SRC["Sourcing: count and type — official, party, expert, citizen, none"]
S --> FM["Form devices: breaking graphic, ticker, music, dramatic replay"]Every variable earns its place: identification variables locate the story in the sample, content variables answer RQ1 and RQ5, sourcing answers RQ2, duration answers RQ3, and form devices answer RQ4. The variable diagram appears in the dissertation's methodology chapter and on the viva table, so the external examiner can audit the path from raw bulletin to every printed table.
Package 01
Builds the constructed two-week sample from published schedules, applies the frozen bulletin-selection rule, and records or downloads each flagged bulletin from official archives with a written archive log (date, source, upload date, duration). Handles the practical failures — missing uploads, delayed bulletins — with pre-fixed substitution rules so the sample stays defensible.
Package 02
Writes the 14-variable codebook with decision rules and worked examples (for instance, when a health story with a political angle counts as politics versus health), pilot-codes 10 percent of the sample, revises ambiguous categories, and freezes version 2. The codebook prints as a dissertation appendix, which is what makes the study replicable.
Package 03
Codes every story in the 42-bulletin sample, then runs the reliability phase: a trained second coder independently re-codes a random 20 percent subsample and Cohen's kappa is computed per variable. Any variable under 0.70 triggers recode-and-retrain before proceeding, and the final kappa table is reported in full — including the weakest variable and what was done about it.
Package 04
Produces the five research-question outputs in SPSS: channel × beat and channel × device chi-square tables with Cramér's V, story-length distributions and ANOVA/Kruskal–Wallis results, mean sources per story, and international-coverage shares. Every table follows BJ-403 conventions — test statistic, degrees of freedom, p-value, effect size — and raw outputs are archived for the viva.
Package 05
Interprets the numbers through gatekeeping theory (Shoemaker and Vos), agenda-setting (McCombs and Shaw), news values (Galtung and Ruge) and Grabe, Zhou and Barnett's sensationalism framework, tying each pattern to the institutional difference between the Prasar Bharati mandate and the BARC-measured private market — without ever claiming to measure intent.
Package 06
Assembles the full dissertation (APA 7 referencing), the replication log — sampling plan, codebook, kappa table, SPSS outputs, archive list — and the viva presentation. The log exists so that the next student can rebuild the study for a different window and extend the series, which is argued in the conclusions chapter.
Slide 1
Public Mandate, Market Pressure — One Evening, Three Channels
Slide 2
Why Bulletins, Not Debates?
Slide 3
Research Questions
Slide 4
Theory Base
Slide 5
Design: Constructed-Week Sampling
Slide 6
The Codebook
Slide 7
Reliability
Slide 8
Findings — What Differed
Slide 9
Interpretation
Slide 10
Ethics and Replication
Slide 11
Conclusion and Future Scope
Bullets, speaker notes and the .pptx · ₹999
Read each answer, close it, and say it in your own words. Your viva will ask. Now you'll know.
Content analysis measures what channels actually broadcast, while editor interviews measure what they say they do — and the two often differ. My research questions are all about output: beats, sources, lengths, devices. The method is also non-reactive, replicable with the same codebook, and teachable under BJ-403, whereas interviews would add self-report bias and approval formalities my 16-week window cannot absorb.
Television content varies systematically by day of the week — weekdays carry heavy political and business coverage, weekends lighten up — so any two consecutive weeks over-represent whatever phase they happened to fall in. A constructed week samples one day of each weekday randomly across the whole 12-week window, so every weekday is represented without dragging in seasonal events. Riffe, Lacy and colleagues showed this design approximates a full census of news content with far less coding.
It is a packaging technique that manufactures urgency independent of the story's substance — in my codebook, the breaking-news graphic, the alert ticker, dramatic music or sting, and dramatic slow-motion replay. The four devices operationalise the formal dimension of Grabe, Zhou and Barnett's 2001 explication of sensationalism, which distinguished content-driven sensational topics from the bells and whistles of form. I coded the devices per story, so prevalence can be compared across channels statistically.
+11 more questions, with answers
Unlocks with the project · ₹999
Unlock for ₹999What to change before it goes anywhere near a submission. Skipping it is how the demo fails live.
Unlocks with the project · ₹999
Unlock for ₹999How to study it, adapt it and walk your examiner through it, step by step.
Unlocks with the project · ₹999
Unlock for ₹999OnlyProjects. (2026). The Public Broadcaster vs Private News: A Comparative Content Analysis of Prime-Time Hindi TV Bulletins: MJMC Broadcast Journalism project bundle [Educational resource]. https://onlyprojects.online/projects/mjmc-broadcast-prime-time-hindi-bulletins-dd-news-comparative-analysis
Keep browsing
Same degree, same syllabus slot, different idea. Pick the one your guide will actually approve.
Print Journalism
Constructed-week sample, two-coder coding sheet, Cohen's kappa and chi-square across TOI, The Hindu, Dainik Jagran and Amar Ujala
PR & Advertising
RACE + PESO campaign with a comparison hostel, Theory of Planned Behaviour survey and a weighed plastic-waste audit