Independent research and analysis project · 2026
Media Attention & High-Profile Crime Case Popularity Analysis
An independent study of why online attention concentrates on some high-profile crime cases and not others, built on an original multi-source dataset covering 11 cases.
The question
Why do some high-profile crime cases attract far more online attention than others, and do documented media events coincide with changes in that attention?
Scope
- 11 high-profile cases, selected against written rules recorded in the source workbook
- Attention measured across three public platforms rather than a single one
- Media-trigger events recorded as dated entries linked back to the cases
- Descriptive and comparative analysis only — no causal claim is made
What I did
- Built the dataset from scratch rather than downloading a finished one.
- Maintained a separate source workbook holding URLs, collection dates, platform metrics, selection rules, and limitations.
- Created consistent Case IDs and standardized date formats so the sheets could be compared to each other.
- Normalized platform-level metrics and created a weighted cross-platform popularity score.
- Built a Media Triggers table linking documented real-world events to the related metrics.
- Revised the original assumptions when the source data did not support them, instead of forcing the data toward the expected conclusion.
Approach
Collecting from the source, not from a shortcut
Each figure in the dataset traces back to an entry in a separate source workbook that records where it came from, the URL, and the date it was collected. That separation matters more than it sounds: platform numbers move, so a metric without a collection date cannot be compared honestly against one gathered weeks later.
Making the sheets comparable
Cases were given consistent Case IDs and dates were standardized to a single format, so a case could be followed across the platform sheets and the Media Triggers table without ambiguity. Platform metrics were then normalized, because raw view counts, Reddit engagement, and search interest are measured on scales that cannot be added together as they stand.
A weighted cross-platform popularity score sits on top of that normalization, giving one comparable figure per case. The exact weights and formulas live in the workbook and are not published here.
Treating triggers as association, not cause
The Media Triggers table records documented events and the metrics around them. Where a rise in attention lines up in time with an event, that is a temporal association and it is described as one. The dataset cannot show that the event caused the attention, and the project does not claim it does.
Changing my mind on the record
The part of this project I would defend most is where the evidence went against what I expected. I revised the assumption rather than the data, and the revision is documented alongside the original.
What exists so far
- Dataset Version 1.0, frozen for analysis
- A separate source workbook with URLs, collection dates, and selection rules
- Standardized Case IDs and dates across sheets
- Normalized platform-level metrics
- A weighted cross-platform popularity score
- A Media Triggers table linking documented events to related metrics
- A written record of assumptions that were revised, and why
No charts or dashboard images are shown yet. The visuals for this project are not finished, and I would rather show nothing than show a mock-up of something I have not built.
What this cannot tell you
The limits are part of the work, not a disclaimer bolted onto it.
- Online attention is not a measure of how serious a case is, how much a victim matters, or whether anyone is guilty. It measures visibility on three platforms and nothing else.
- A change in attention that coincides with a documented event is a temporal association. This dataset cannot establish that the event caused the change.
- Each platform measures something different — views, engagement, and relative search interest — so any combined score is an interpretation, not a natural quantity.
- Platform metrics move over time and are shaped by recommendation systems, so the figures reflect the collection dates recorded in the workbook rather than a fixed truth.
- Eleven cases is a small, deliberately selected set. Findings describe those cases and do not generalize to coverage as a whole.
- Availability differs by case and by platform, so absence of data is not evidence of absence of attention.
What I learned
- Writing down the selection rules before collecting made it much harder to quietly change the criteria later.
- Normalization is a decision, not a formality — the choice of scale changes which cases look prominent.
- Recording the collection date turned out to be as important as recording the number itself.
- Distinguishing what the data shows from what I expected it to show is the discipline the project is really about.
Next steps
Planned work. None of this is finished yet.
- Complete the Power BI visualization phase: comparing platform attention, plotting media triggers against metrics over time, and showing differences between cases.
- Write a short methodology note that can be published alongside the dashboard.
- Publish an executive summary once the visuals support it.