# Project Description
Following is the details of the final project for MC4AI class April-June 2023. The project is created by Mai Ngọc Lan Vy
# About the Dataset
This is a survey dataset of 9,610 participant across the United States regarding their experience with financial services.

Survey participants are individuals who are responsible for, or share in, making household financial decisions.

Sample of the dataset can be found [here](https://docs.google.com/spreadsheets/d/1BHP8J1cIRFYdFwOcRVq539XVueTEr9mhPHTui25Akic/edit?usp=sharing).

# Variables
The dataset has:
* 3 housekeeping variables: record number, participant identifier, and completion time and date

| record | uuid | date |
| -------- | -------- | -------- |
| 1 | hrmxghhs2vqf6wvu | 12/14/21 14:30 |
| 2 | bhe6t05ta0sp5xaf | 12/14/21 14:23 |
| 5 | p3y1s822xxk2rk1u | 12/14/21 14:37 |

* 35 demographic variables (e.g: race, language spoken, etc.)

| S2 |...| D1r2 |...| D6 |...|
| -- | - | ---- |---| -- | - |
| 2 |...| 0 |...| 1 |...|
| 1 |...| 1 |...| 1 |...|
| 2 |...| 0 |...| 1 |...|

* 266 questionaire response variables (e.g: what financial products do you have, etc.)

| Q1r1 |...| Q6r16oe |...| Q11o2 |...| Q24 |
| ---- | - | ------------ |---| ----- | - | --- |
| 0 |...|They had ATMs!|...| |...| 4 |
| 0 |...| |...| |...| 6 |
| 0 |...| |...| auto |...| 6 |

* 7 variables classifying each response to a segment group (e.g: generation, income, etc.)

| GENERATION | INCOME | RACE |
| ---------- | ------ | ----- |
| 4 | 3 | 1 |
| 2 | 3 | 4 |
| 4 | 2 | 1 |

* 1 weight variable

| WEIGHT |
| ----------- |
| 3.37279615 |
| 0.415357658 |
| 3.746498048 |

> All variables and row responses are coded and should be accompanied with the [data dictionary](https://docs.google.com/spreadsheets/d/1B_bxqnDFLVssaK124etqCnW-uSHQ7_fU/edit?usp=sharing&ouid=102529998437857684808&rtpof=true&sd=true
).

# Type of Variable
All variables are categorical with
1) a value of 0 (no) or 1 (yes) to the questionaire option
2) ordinal / rank / Linkert scale with a range of 1 to 7, or a range of 1 to 3
3) a free response
4) a number indicate a classified segment (e.g: the GENERATION variable above)

Some example of the type of variables

| Q3 | Q18r1 | Q25r1 |
| -------- | -------- | -------- |
| 7 | 5 | 1 |
| 7 | 1 | 2 |
| 6 | 4 | 3 |

#### Q3 | Overall, How satisfied are you with your primary financial institution?
1 = Not All All Satisfied
...
7 = Extremely Satisfied
#### Q18r1 | *Speed of lending me the money I need* -- How important are the following when you consider taking out a loan?
1 = Not All All Important
...
7 = Extremely Important
#### Q25r1 | *Financial Well-Being* -- Please rank the top 3 causes or movements that matter to you most from the list below.
1 = Most important
2 = Second most important
3 = Third most important

# Data Analysis Ideas
1. Compare participant considering credit union as their primary financial institution (PFI) vs. non-credit union insitution, across all questionaires (e.g: satisfaction, trust, product ownership -- loan, insurance, investment, etc.). What does credit union do better/worse compare to bank and other financial insitutions?
2. Identify areas of improvement (in terms of services) and growth (in terms of product) for credit union.
3. Who/What demographic are most likely to bank with credit union and why?
4. Provide prediction for a participant willingness to purchase a financial product in the next 5 year if they already have the 1, 2 or more of the same or relevant products? How does that differ for each racial, age/generation, income, and LGBTQ+ category?
5. What worries predict one's current financial product, or willingness to purchase a financial product in the next 5 year, or willingness to work with a financial advisor?
6. Identify the best marketing channels for each racial, age/generation, income, and LGBTQ+ category.
7. How satisfaction with DEI aspects at PFI affect the amount of product currently have or to be purchase in the next 5 year at PFI
8. Idenfify all factors affect Net Promoter Score/satisfaction score/trustworthiness score. Classify based on demographic segments.