Bedtime Smartphone Use & Sleep Efficiency:
Blue Light Filter as a Moderating Factor

The FIRST correlational study with OBJECTIVE measures for both sleep quality and smartphone use!


This project was completed as my 4th year undergraduate directed studies project, extending upon the UBC Smartphone and Sleep Study by UBC Social Health Lab.

The UBC Smartphone & Sleep Study recruits people of ages 14+ from all over Metro Vancouver, which gave me (as the former co-study coordinator) the opportunity to interview people from a wide variety of demographics.

After many conversations with the participants upon completion of their week-long participation, I noticed a few people mentioning their perceived benefits of using the blue-light filter feature on their smartphones, whereas some participants weren’t even aware of this feature on their smartphones. This sparked my interest towards comparing these groups, and led me to specifically focus on the potential moderating factor of blue light filter use in the relationship between bedtime smartphone social media use and sleep quality (both objective & subjective measures).

The full directed studies paper can be found in the below PDF file:

For those who don’t want to read the paper


Many studies on the topic conclude that bedtime phone use is related to negative sleep outcomes in both adult and adolescent samples. In adults, bedtime phone use after lights out, as well as longer bedtime smartphone screen-time significantly predicted negative sleep outcomes – specifically worse sleep quality (i.e. worse sleep efficiency, more sleep disturbance, more daytime dysfunction). Smartphone ownership was related to more bedtime media use, later bedtimes, shorter sleep duration, and more sleeping problems in adolescents. Greater social media use, especially around bedtime, has also been shown to be significantly linearly associated with poor sleep.

Problem With Previous Research

One major flaw in these past studies is that almost all of them solely rely on self-reported phone use and sleep quality. In order to properly investigate what factors contribute to this relationship, it is crucial for future researchers to use both subjective and objective measures.

A potential contributing factor that has been discussed is the blue light that emits from the LED screens of our smartphones, which can suppress the production of melatonin, a hormone regulating our sleep-wake cycle. Therefore, using a blue-light filter could potentially mitigate the negative sleep outcomes predicted with bedtime smartphone use.

Research Question

How does objectively measured smartphone use 1 hr before sleep relate to subjective and objective measures of sleep quality & does blue light filter use serve as a moderating factor?


All participants were recruited through the UBC Smartphone & Sleep Study website, Facebook and Instagram advertisements, UBC Human Subjects Pool, as well as posters around the city of Metro Vancouver. People who were interested in participating in the study contacted research assistants to schedule the first initial session to get set up with the study materials and instructions. At the first 30-40 minute meeting with the researcher, research assistants verbally explained all the tasks involved during the 7-night participation of the study. After being fully informed of all aspects of the study, the researchers obtain verbal and electronic consent before setting up individual participants with all the study materials. Participants are then asked to download ‘Betrack’, an Android research app that tracks the duration and frequency of each individual app usage on their smartphones. The participants are also provided with a movement tracking wristband ‘Actiwatch’ for 7 nights of sleep, as well as a sleep diary. With the Actiwatch, the participants are asked to set a time marker on the wristband by pressing a button when they attempt to go to sleep (i.e. bedtime onset), as well as when they finish attempting to sleep. The Actiwatch system can also predict bedtime onset & wake-up time with the movement activity data, in case the participants forget to indicate the time marker. An online demographics survey is administered at both the first and final meeting with the researcher.


The primary predictor variables being examined in this study are the duration of daily smartphone app activity 1 hour before sleep, as well as the use of blue light filter as a moderating factor. The main outcome variable analyzed will be the day-to-day objective sleep efficiency score from the accelerometer. The participants are also given a “sleep diary” with a set of validated questions to determine any unexpected circumstances that may have influenced their nightly sleep. Duration and frequency of smartphone use is objectively measured with an Android smartphone app-tracking application, ‘BeTrack’. Participants’ use of blue light filter is obtained with a question in the self-report demographics survey during the second session. Sleep efficiency scores are obtained from the movement tracking wristband ‘Actiwatch’, and calculated as total duration of sleep divided by the total duration of time in bed on that same night.

Objective Sleep Quality

Sleep efficiency scores obtained from an accelerometer wristband (i.e. Actiwatch Spectrum)

Subjective Sleep Quality

Consensus Sleep Diary (Carney et al., 2012) 

Pittsburg Sleep Quality Index (PSQI; Buysse et al., 1989)

Blue Light Filter Use

Self-reported through questionnaire

Bedtime Smartphone Use

BeTrack research app (i.e. frequency on type of apps & duration of use)


Consistent with past research, it will be hypothesized that objectively measured sleep efficiency scores derived from the Actiwatch will be negatively correlated with the duration of smartphone social media activity 1 hour before bedtime (i.e. higher the social media use, lower the sleep quality & vice versa). Blue light filter use will also be hypothesized to moderate the strength of the relationship between sleep efficiency (both objective and subjective) and bedtime smartphone use.

The adolescent sample and the adult sample will also be divided for analysis, due to potential differences in lifestyle, sleep patterns, and developmental differences.

Future Data Analysis

After data collection is complete, moderated multiple regression will be used to analyze whether the sleep efficiency score (i.e. interval outcome variable) objectively measured from the Actiwatch will be negatively correlated with the duration of smartphone activity 1 hour before sleep attempt (i.e. continuous predictor variable). Blue light filter will be examined as the dichotomous moderator variable for the regression data analysis.

How does personal space invasion affect behaviour and attention?

Learn more about my work at UBC’s Brain, Attention & Reality Lab.