# Understanding US Congressional Election Polls and Prediction Models

---

#### Keshif built an exploratory dashboard to analyze and geo-spatially visualize a rich proprietary data model based on multilevel regression with poststratification (MRP) to predict Congressional race outcomes and potential upsets for the 2018 House of Representatives elections.

|     |     |
| --- | --- |
| Client |
| Region<br>North America | Year<br>2018 |
| Sectors<br>- Democracy and Governance | Solutions<br>- Election Results<br>- Surveys |

# Impact

The dataset derives from monthly polling of 50,000 households conducted by HarrisX, a nationally recognized polling agency in the United States. Using our solution, audiences can explore results for 435 Congressional Districts through interactive regional and hexagonal maps, and a detailed tabular district view consisting of information related to district status, forecasts, the incumbent party, and approval ratings of the U.S. President.

## Services

##### Visual Data Analytics and Dashboarding

##### Web Design and Development