FitBit Fitness Tracker Data Analysis

Author: Mohammed Amja

Date: 18/08/2021

Note: Please refer to the files in this repository to find all of the data/code/graphs/tables found in this report and much more.

Tableau Visualizations: Click here

---

Introduction

Bellabeat is a high-tech company that manufactures health-focused products primarily for women. The company makes products and applications that help monitor their health and wellness such as activity, stress levels, sleep and other mindful habits. Bellabeat is a successful small company that is looking forward to expand their capabilities and improve their products to hopefully become a competitive large player in the global smart device market. Urška Sršen, cofounder and Chief Creative Officer of Bellabeat, is determined to unlock new growth oppurtunities and believes that can be done by analyzing data from smart device fitness gadgets.

---

Problem Statements

The Co-founder and Chief Creative Officer of Bellabeat, Urška, wants to develop effective marketing strategies to help unlock new growth oppurtunities for the company. She believes this can be done by analyzing consumer data on smart device usage and gaining insights.

The Business Task:

• To analyze data to gain insights on how consumers use smart devices.
• Use the discovered insights to help guide marketing strategies for the company.

---

Preparing Data for Analysis

For this project, I've used the 18 datasets dated from 03/12/2016 - 05/12/2016 provided on kaggle. Click on this FitBit Fitness Tracker Data (CC0: Public Domain, dataset made available through Mobius) to access the website and download the datasets provided as .csv files. This data set contains personal fitness data from 30 fitbit users on minute-level output for physical activity, heart rate, and sleep monitoring. It also includes information on daily activity, steps, and heart rate that can be used to explore users’ habits. Thirty Fitbit users consented to the submission of personal tracker data and the data provided in this website is made available to access to the public.

Or you could access and download the data from this repository named as "Raw Data".

I am using Microsoft SQL Server Management Studio in this part of the project to help process and analyze the datasets.

First make sure to import all of the dataset as a .csv file to the database server. In order to solve this business task, only 6 of the given 18 datasets was used. Many of the files are either redundant, is not essential, or relevant, or lacking sufficient data to perform an analysis upon.

---

Processing of Data

Here, I will be transforming and organizing data by adding new columns, extracting information and removing bad data and duplicates.

In order to get accurate analysis, validate and make sure the dataset does not include any bias, incorrect data, and duplicates.

--Checking Number of Rows on dailyActivities

Select Count(*)
From [dbo].[dailyActivity_merged]


--Checking for duplicates in dailyActivity dataset

Select Id, ActivityDate, TotalSteps, Count(*)
From [dbo].[dailyActivity_merged]
group by id, ActivityDate, TotalSteps
Having Count(*) > 1


--Modify date format for better understaning in sleepDay

Update sleepDay_merged
Set SleepDay = Convert(date, SleepDay, 101)


--Modify date format for better understaning in dailyActivity

Update dailyActivity_merged
Set ActivityDate = Convert(date, ActivityDate, 101)


--Add day_0f_week column on dailyActivities

Alter Table [dbo].[dailyActivity_merged]
ADD day_of_week nvarchar(50)


--Extract datename from ActivityDate

Update dailyActivity_merged
SET day_of_week = DATENAME(DW, ActivityDate)


--Add sleep data columns on dailyActivity table

Alter Table [dbo].[dailyActivity_merged]
ADD total_mins_sleep int,
total_mins_bed int


--Add sleep records into dailyActivity table

UPDATE dailyActivity_merged
Set total_mins_sleep = t2.TotalMinutesAsleep,
total_mins_bed = t2.TotalTimeInBed 
From [dbo].[dailyActivity_merged] as t1
Full Outer Join sleepDay_merged as t2
on t1.id = t2.id and t1.ActivityDate = t2.SleepDay


--Adding specific date format to [dailyActivity_merged] table

Alter table dailyActivity_merged
Add Date_d date

Update dailyActivity_merged
Set Date_d = CONVERT( date, ActivityDate, 103 )


--Split date and time seperately for [hourlyCalories_merged] table

Alter Table [dbo].[hourlyCalories_merged]
ADD time_h int

Update [dbo].[hourlyCalories_merged]
Set time_h = DATEPART(hh, Date_d)

Update [dbo].[hourlyCalories_merged]
Set Date_d = SUBSTRING(Date_d, 1, 9)


--Split date and time seperately for [hourlyIntensities_merged]

Alter Table [dbo].[hourlyIntensities_merged]
ADD time_h int

Update [dbo].[hourlyIntensities_merged]
Set time_h = DATEPART(hh, ActivityHour)

Update [dbo].[hourlyIntensities_merged]
Set ActivityHour = SUBSTRING(ActivityHour, 1, 9)


--Split date and time seperately for [hourlySteps_merged]

Alter Table [dbo].[hourlySteps_merged]
ADD time_h int

Update [dbo].[hourlySteps_merged]
Set time_h = DATEPART(hh, Date_d)

Update [dbo].[hourlySteps_merged]
Set Date_d = SUBSTRING(Date_d, 1, 9)


--Split date and time seperately for [minuteMETsNarrow_merged]

Alter Table [dbo].[minuteMETsNarrow_merged]
ADD time_t time

Update [dbo].[minuteMETsNarrow_merged]
Set time_t = CAST(Date_d as time)

Update [dbo].[minuteMETsNarrow_merged]
Set time_t = Convert(varchar(5), time_t, 108)

Update [dbo].[minuteMETsNarrow_merged]
Set Date_d = SUBSTRING(Date_d, 1, 9)


--Create new table to merge hourlyCalories, hourlyIntensities, and hourlySteps

Create table hourly_cal_int_step_merge(
Id numeric(18,0),
Date_d nvarchar(50),
time_h int,
Calories numeric(18,0),
TotalIntensity numeric(18,0),
AverageIntensity float,
StepTotal numeric (18,0)
)


--Insert corresponsing data and merge multiple table into one table

Insert Into hourly_cal_int_step_merge (Id, Date_d, time_h, Calories, TotalIntensity, AverageIntensity, StepTotal)
(Select t1.Id, t1.Date_d, t1.time_h, t1.Calories, t2.TotalIntensity, t2.AverageIntensity, t3.StepTotal
From [dbo].[hourlyCalories_merged] as t1
Inner Join [dbo].[hourlyIntensities_merged] as t2
ON t1.Id = t2.Id and t1.Date_d = t2.ActivityHour and t1.time_h = t2.time_h
Inner Join [dbo].[hourlySteps_merged] as t3
ON t1.Id = t3.Id and t1.Date_d = t3.Date_d and t1.time_h = t3.time_h)


--Checking for duplicates

/*Select Id, time_h, Calories, TotalIntensity, AverageIntensity, StepTotal, Count(*) as duplicates
From [dbo].[hourly_cal_int_step_merge]
Group by Id, time_h, Calories, TotalIntensity, AverageIntensity, StepTotal
Having Count(*) > 1*/


--Checking for duplicates

/*Select sum(duplicates) as sum_s
from (Select Id, Date_d time_h, Calories, TotalIntensity, AverageIntensity, StepTotal, Count(*) as duplicates
From [dbo].[hourly_cal_int_step_merge]
Group by Id, Date_d, time_h, Calories, TotalIntensity, AverageIntensity, StepTotal
Having Count(*) > 1
Order by duplicates DESC) as cte*/


--Query in hh:mm time format for better understanding on MET Table 

select Id, Cast(Date_d as date) as date_d, METs, Convert(varchar(5), time_t, 108) as time_t
From [dbo].[minuteMETsNarrow_merged]


--Change date type nvarchar to date on MET table to join properly with other table

Alter table minuteMETsNarrow_merged
Add dates_d date

Update minuteMETsNarrow_merged
Set dates_d = Cast(Date_d as date)

---

Analysis of Data

Here, I will be analysing the consumer data to discover trends and patterns.

                                                      --Analysis--
--Calculate average met per day per user, and compare with the calories burned

Select Distinct t1.Id, t1.dates_d, sum(t1.METs) as sum_mets, t2.Calories
From [dbo].[minuteMETsNarrow_merged] as t1
inner join dailyActivity_merged as t2
on t1.Id = t2.Id and t1.dates_d = t2.Date_d
Group By t1.Id, t1.dates_d, t2.Calories
Order by dates_d

Result: Data contains 934 rows, this table limits to view the top 50 rows

Id	dates_d	sum_METs	Calories
1503960366	12-04-2016	25241	1985
1624580081	12-04-2016	17234	1432
1644430081	12-04-2016	22768	3199
1844505072	12-04-2016	21704	2030
1927972279	12-04-2016	15599	2220
2022484408	12-04-2016	23035	2390
2026352035	12-04-2016	18830	1459
2320127002	12-04-2016	23090	2124
2347167796	12-04-2016	24924	2344
2873212765	12-04-2016	22570	1982
3372868164	12-04-2016	19118	1788
3977333714	12-04-2016	20644	1450
4020332650	12-04-2016	26521	3654
4057192912	12-04-2016	18434	2286
4319703577	12-04-2016	14400	2115
4388161847	12-04-2016	14400	2955
4445114986	12-04-2016	18392	2113
4558609924	12-04-2016	20576	1909
4702921684	12-04-2016	21040	2947
5553957443	12-04-2016	22118	2026
5577150313	12-04-2016	26431	3405
6117666160	12-04-2016	14400	1496
6290855005	12-04-2016	17890	2560
6775888955	12-04-2016	14400	1841
6962181067	12-04-2016	22065	1994
7007744171	12-04-2016	27171	2937
7086361926	12-04-2016	24496	2772
8053475328	12-04-2016	26168	3186
8253242879	12-04-2016	20620	2044
8378563200	12-04-2016	24210	3635
8583815059	12-04-2016	18896	2650
8792009665	12-04-2016	17433	2044
8877689391	12-04-2016	32010	3921
1503960366	13-04-2016	22859	1797
1624580081	13-04-2016	16990	1411
1644430081	13-04-2016	20656	2902
1844505072	13-04-2016	19880	1860
1927972279	13-04-2016	15014	2151
2022484408	13-04-2016	25074	2601
2026352035	13-04-2016	19636	1521
2320127002	13-04-2016	21782	2003
2347167796	13-04-2016	21675	2038
2873212765	13-04-2016	22820	2004
3372868164	13-04-2016	22380	2093
3977333714	13-04-2016	21286	1495
4020332650	13-04-2016	14400	1981
4057192912	13-04-2016	18566	2306
4319703577	13-04-2016	17299	2135
4388161847	13-04-2016	19246	3092
4445114986	13-04-2016	18230	2095

---

                                                             
--Activities and colories comparison

Select Id,
SUM(TotalSteps) as total_steps,
SUM(VeryActiveMinutes) as total_Vactive_mins,
Sum(FairlyActiveMinutes) as total_Factive_mins,
SUM(LightlyActiveMinutes) as total_Lactive_mins,
SUM(Calories) as total_calories
From [dbo].[dailyActivity_merged]
Group By Id

Result: Strong correlation between activity intense time and calories burned

id	total_steps	total_Vactive_m	total_Factive_m	total_Lactive_m	calories
8583815059	223154	300	688	4287	84693
6117666160	197308	44	57	8074	63312
8053475328	457662	2640	297	4680	91320
4319703577	225334	111	382	7092	63168
4388161847	335232	718	631	7110	95910
6290855005	163837	80	110	6596	75389
2320127002	146223	42	80	6144	53449
1644430081	218489	287	641	5354	84339
2873212765	234229	437	190	9548	59426
1503960366	375619	1200	594	6818	56309
4702921684	265734	159	807	7362	91932
4445114986	148693	205	54	6482	67772
4057192912	15352	3	6	412	7895
1624580081	178061	269	180	4758	45984
8792009665	53758	28	117	2662	56907
2026352035	172573	3	8	7956	47760
3977333714	329537	567	1838	5243	45410
5577150313	249133	2620	895	4438	100789
4558609924	238239	322	425	8834	63031
8378563200	270249	1819	318	4839	106534
1844505072	79982	4	40	3579	48778
7007744171	294409	807	423	7299	66144
8253242879	123161	390	272	2221	33972
2347167796	171354	243	370	4545	36782
4020332650	70284	161	166	2385	73960
6962181067	303639	707	574	7620	61443
6775888955	65512	286	385	1044	55426
5553957443	266990	726	403	6392	58146
1927972279	28400	41	24	1196	67357
7086361926	290525	1320	786	4459	79557
3372868164	137233	183	82	6558	38662
8877689391	497241	2048	308	7276	106028
2022484408	352490	1125	600	7981	77809

---

--Average Sleep Time per user

Select Id, Avg(TotalMinutesAsleep)/60 as avg_sleep_time_h,
Avg(TotalTimeInBed)/60 as avg_time_bed_h,
AVG(TotalTimeInBed - TotalMinutesAsleep) as wasted_bed_time_m
from sleepDay_merged
Group by Id

Result:

Id	avg_sleep_time_h	avg_time_bed_h	wasted_bed_time_m
1503960366	6.004666	6.386666	22.92
1644430081	4.9	5.766666	52
1844505072	10.866666	16.016666	309
1927972279	6.95	7.296666	20.8
2026352035	8.436309	8.960714	31.464285
2320127002	1.016666	1.15	8
2347167796	7.446666	8.188888	44.533333
3977333714	4.894047	7.685714	167.5
4020332650	5.822916	6.329166	30.375
4319703577	7.94423	8.366025	25.307692
4388161847	6.71875	7.103472	23.083333

---

--Sleep and calories comparison	

Select t1.Id, SUM(TotalMinutesAsleep) as total_sleep_m,
SUM(TotalTimeInBed) as total_time_inbed_m,
SUM(Calories) as calories
From [dbo].[dailyActivity_merged] as t1
Inner Join [dbo].[sleepDay_merged] as t2
ON t1.Id = t2.Id and t1.ActivityDate = t2.SleepDay
Group By t1.Id

Result:

Id	total_sleep_m	total_time_inbed_m	calories
1503960366	9007	9580	46807
1644430081	1176	1384	11911
1844505072	1956	2883	5029
1927972279	2085	2189	11581
2026352035	14173	15054	43142
2320127002	61	69	1804
2347167796	6702	7370	29570
3977333714	8222	12912	43691
4020332650	2795	3038	25560
4319703577	12393	13051	52642
4388161847	9675	10229	75159
4445114986	10785	11671	61128
4558609924	638	700	10985
4702921684	11792	12375	85211
5553957443	14368	15682	58146
5577150313	11232	11976	92019
6117666160	8618	9183	44295
6775888955	1049	1107	7034
6962181067	13888	14450	61443
7007744171	137	143	4301
7086361926	10875	11194	63783
8053475328	891	905	9928
8378563200	14187	15466	110539
8792009665	6535	6807	34490

---

--Daily Sum Analysis - No trends/patterns found

Select SUM(TotalSteps) as total_steps,
SUM(TotalDistance) as total_dist,
SUM(Calories) as total_calories,
day_of_week
From [dbo].[dailyActivity_merged]
Group By  day_of_week

Result: No trends or patterns found

total_steps	total_dist	total_calories	day_of_week
1133906	763	345393	Wednesday
1010969	676	292016	Saturday
933704	613	278905	Monday
838921	554	273823	Sunday
938477	614	293805	Friday
1088658	718	323337	Thursday
1235001	817	358114	Tuesday

---

--Daily Average analysis - No trends/patterns found

Select AVG(TotalSteps) as avg_steps,
AVG(TotalDistance) as avg_dist,
AVG(Calories) as avg_calories,
day_of_week
From [dbo].[dailyActivity_merged]
Group By  day_of_week

Result: No trends or patterns found

avg_steps	avg_dist	avg_calories	day_of_week
7559.373333	5.086666	2302.62	Wednesday
8152.975806	5.451612	2354.967741	Saturday
7780.866666	5.108333	2324.208333	Monday
6933.231404	4.578512	2263	Sunday
7448.230158	4.873015	2331.785714	Friday
7405.836734	4.884353	2199.571428	Thursday
8125.006578	5.375	2356.013157	Tuesday

---

--Time Expenditure per day
Select Distinct Id, SUM(SedentaryMinutes) as sedentary_mins,
SUM(LightlyActiveMinutes) as lightly_mins,
SUM(FairlyActiveMinutes) as fairlyactive_mins, 
SUM(VeryActiveMinutes) as veryactive_mins
From [dbo].[dailyActivity_merged]
where total_mins_bed IS NOT NULL
Group by Id

Result:

---

Visualizing Data

In this phase, we will be visualizing the data analyzed and tables created using Tableau Public.

For the interactive version, Click here

Activity Durations and Calories Relationship:

Key Findings:

• The R-Squared value for Low Active graph is 0.0118
• The R-Squared value for Fairly Active graph is 0.0391
• The R-Squared value for Very Active graph is 0.3865

There is a strong correlation between Very Active minutes and the amount of calories burned. The r-squared value seems to rise as the intensity and the duration is the activity increases.

Thus by inferring to the r-squared values of the respective trend lines of the graphs, we can conclude that the higher the intensity and the duration of the activity, the more calories is burned.

METs and Average Calories Burned:

A brief overview on METs:

Metabolic Equivalent of Task (MET) is the ratio of your working metabolic rate relative to your resting metobolic rate. It is a way to describe the intensity of an activity or workout. METs provide a simple and practical way to easily understand the body's energy expenditure with metabolic equivalents.

One MET is defined as resting metabolic rate, it is approximately 3.5 milliliters of oxygen consumed per kilogram of body weight per minute. One MET is the energy a person spends while resting or sitting idle.

So an MET of 4 means a person is spending 4 times the energy they are when sitting idle. The METs for each person is not constant and can have slight variations depending on their weight, age or other physical parameters.

To calculate the amount of calories burned per minute, we can use the formula:

Calories burned per minute = (METs x 3.5 x (your body weight in Kg)) / 200

To better understand METs more, the approximate METs values while performing certain excersices are depicted below:

Light < 3.0 METs	Moderate 3.0–6.0 METs	Vigorous > 6.0 METs
Sitting at a desk: 1.3	Housework (cleaning, sweeping): 3.5	Walking at very brisk pace (4.5 mph): 6.3
Sitting, playing cards: 1.5	Weight training (lighter weights): 3.5	Bicycling 12–14 mph (flat terrain): 8
Standing at a desk: 1.8	Golf (walking, pulling clubs): 4.3	Circuit training (minimal rest): 8
Strolling at a slow pace: 2.0	Brisk walking (3.5–4 mph): 5	Singles tennis: 8
Washing dishes: 2.2	Weight training (heavier weights): 5	Shoveling, digging ditches: 8.5
Hatha yoga: 2.5	Yard work (mowing, moderate effort): 5	Competitive soccer: 10
Fishing (sitting): 2.5	Swimming laps (leisurely pace): 6	Running (7 mph): 11.5

Key Findings:

• The R-Squared value is 0.5504
• Strong positive corellation between METs and average calories burned.

The amount of calories burned for every user is highly dependent on their MET values they spend every day. This can be seen by the high r-squared value suggesting that the trend line has strong relation with the data points.

**Sleep and Calories Comparison **

Key Findings:

• The R-Squared value is 0.8727
• Strong positive corellation between amount of sleep and calories burned.

Higher duration of sleep is associated with higher amount of calories burned. An adequate duration and good quality of sleep constitutes to higher calories burned during the sleeping process.

However sleeping more than the required range doesn't seem to burn more calories and in fact causes the opposite to occur, which is burn fewer calories.

Popular Time for Activities

From the graph above, we can infer that the most popular time people are active throughout the day is between 5:00 AM - 9:00PM

Tableau Dashboard

Visualizations built in a dashboard.

---

Conclusion

After performing the collection, transformation, cleaning, organisation and analysis of the given datasets, we have enough factual evidence to suggest answers to the business-related questions that were asked.

We can infer that the duration and the level of intensity of the activities performed are greatly in dependence to the amount of calories burned. METs provide a great insight on the intensity of activities performed and the amount of calories burned per minute. While most of the consumers attain adequate amounts of sleep, it is noticed that a small fraction of the users either oversleep or undersleep. Consumers are also more likely to perform low-high intensity activities in the range of 5:00 AM - 9:00PM throughout the day.

In order to design new marketing strategies to better focus on unlocking new growth oppurtunities and develop the business, we have to refer to the analysis provided above and keep those facts in mind. The recommendations I would provide to help solve this business-related scenario is shown below.

Top Recommendations to Marketing Strategists:

• Highlight the MET tracking feature on the smart devices as a marketing strategy and create awareness on MET values. For it allows users to track their level of intensity of activities and provide a real time insight on how much calories they burn every minute.
• Consumers seem to spend most of their time inactive and live a sedentary lifestyle. Notifying users through smart device notifications during the most popular time for performing activities which is between 5:00 AM - 9:00PM can remind people to exercise and live a more active lifestyle.
• Provide app notification for users to remind them to get sufficient sleep every day and implement new sleep measurement features or products such as tracking Rapid Eye Movement (REM) sleep.
• Consider setting daily/weekly calorie challenges and award points to users based on the top performers. Where the points can be accumulated and redeemed as a discount for their next product purchase.