<Home

Practical
Recommendations

In this section of the EUPASMOS toolkit it is described the main practical recommendations that will help guiding the implementation of the PA and sport monitoring system in your country, region or local community.

Each step of the process is presented in an at-glance approach to ensure readiness for data collection and analysis, supporting the development of more adjusted policies and actions in Sport and Physical Activity promotion at different governance levels.

Planning

Planning is one of the most relevant phases when preparing the implementation of a PA and sport monitoring survey. The main issues that should be taken into consideration are as follows:

  1. Read carefully the EUPASMOS Toolkit and Manual of Operations;
  2. Make sure you have all the necessary materials to implement the survey (validated questionnaires, equipment’s for anthropometric data collection, and accelerometers);
  3. If questionnaires are not validated in your country language, develop a validation process as described in the EUPASMOS Manual of Operations;
  4. Contact a wide range of institutions to ensure the largest possible representative sample for your survey; use different communication platforms to increase project dissemination;
  5. At national level surveys, aim to include a sample of, at least, 1000 individuals;
    1. Sample distribution should be balanced according to the country population distribution:
      1. Sex (e.g. 50% men/50% women);
      2. Age groups (18-34, 35-49, 50-64, >65 years-old);
      3. Socio-economic status;
      4. Pregnant women and people with locomotor disabilities;
  6. For smaller and specific population surveys, aim also to enrol a balanced sample distribution accordingly;
  7. Create an objective plan and a concrete timeline to collect data;
  8. Train all team members regarding the methodology procedures;
  9. On the first day of data collection ensure that all the measurement instruments are available;
  10. Accelerometer delivery to each participant should be done personally by a trained survey team member;
  11. After the 7-days monitoring period (using the accelerometer), ensure all the necessary conditions to receive the equipment’s and download the data.
  12. If necessary, additional questionnaires can be also applied during this second contact moment with the survey team;
  13. Ensure that all participants receive the reports resulting from data collection and analysis timely.
DOWNLOAD
EUPASMOS Informed consent

Questionnaires

  • Check if there is a translated version of the questionnaires in your country language;
  • If yes, download the version to be administered during the survey;
  • If it is not translated, the survey team should find a specialist to translate the questionnaire;
  • After the questionnaire translation, the team must run a test-retest procedure using, at least, 20 participants; the procedures are described in the EUPASMOS Manual of Operations;
  • After the test-retest validation, questionnaire is ready to be used according to the EUPASMOS framework.
Questionnaires used by EUPASMOS to measure PA and sport participation: Questionnaires used by EUPASMOS to measure PA and sport correlates and determinants: Additional questionnaires can be applied to assess health status and socio-demographic information from the participants in the survey. For these purposes, EUPASMOS framework used the EHIS questionnaire (Wave 3), which is the instrument used by the EUROSTAT at European level to collect this information. Both questionnaires are presented in the EUPASMOS Manual of Operations.
DOWNLOAD

EUPASMOS questionnaires validated in partner’s language

Accelerometry

To objectively measure physical activity levels, EUPASMOS used the UKK RM42 triaxial accelerometer along with a new method to process the raw data collected. Other validated models of triaxial accelerometers, using the MAD-APE algorithms to process raw data can be used.

Legend: UKK RM42 triaxial accelerometer.

Main steps and procedures to be followed:
  • Deliver, personally, the accelerometer to the participant (preferably);
  • Give all the instructions to the participant about how to wear the device during awake and sleep time (24-h, except for water based activities, during 7 consecutive days, including 2 weekend days);
  • Match participant codification with the accelerometer device code;
  • Identify participant responsibilities and tasks;
  • Preferably, receive the accelerometer directly from the participant after the measurement period.

Participants are required to record, in an accelerometer idle dairy, all occasions that they removed the device and identify the reason. A template specific form is available from EUPASMOS framework.

Whenever possible, it is suggested to complete the information about the physical activity developed and sport participation by asking participants to fill out a physical activity diary. This diary will help to discriminate more precisely the activities performed during the period of data collection.

DOWNLOAD

Analysis

Within the EUPASMOS framework, it is suggested to develop specific procedures for data analysis.

Questionnaires
  • If the questionnaires used to collect data in your country are not translated and or validated, a test-retest procedure, as described in EUPASMOS Manual of Operations should be performed previously to data collection (using, at least, 20 participants);
    • Statistical procedures to develop the assessment of questionnaire validation should be developed as described in the referred Manual;
      • Acceptability, exploratory factorial analysis, internal consistency, reproducibility, and convergent validity are some of the main procedures that can be developed.
  • If the questionnaires used to collect data in your country are already translated and or validated, the survey team can proceed with data collection.
  • After data collection, databases should be filled. In order to ensure the best possible quality of data, any database should be cleaned from errors and outliers.
    • EUPASMOS developed a specific software tool to analyse databases and identify errors and outliers;
    • Develop the necessary cleaning procedures in your databases.
  • After cleaning the databases resulting from questionnaires application, several analysis procedures can be performed:
    • Control for normality and homoscedasticity of the study variables;
    • When necessary, log transformations to normalize distributions can be used;
    • Calculate means ± standard deviations;
    • Multiple linear regressions models, adjusted for confounding variables, should be performed to study the independent associations between key variables like age, sex, BMI and other socio-demographic or health markers with relevant physical activity, sport and sedentary behaviour variables;
      • In order to facilitate the comparisons of the results obtained in the multiple linear regression models, standardized beta coefficients should be presented;
      • To determine how much the independent variables are linearly related to one another, the multicollinearity should be studied by statistic tolerance (1-R 2 ), being the stability of the regression model disturbed by multicollinearity if tolerance is inferior to 0.1.
      • Statistical significance should be set at P < 0.05.
Accelerometry

EUPASMOS used the mean amplitude deviation (MAD) method to process the accelerometer raw data, which can be applied to different accelerometer brands and seems to present high validity for estimating all physical activity intensities including the different types of sedentary behaviours.

For harmonization reasons, the accelerometer data processing was performed by automatic algorithms that were already included in the EUPASMOS cloud provided by UKK (Finnish project partner). Each country's data were uploaded to this cloud and, after an automatic process, the files containing the raw data were shared with each survey partner and then with the project coordination centre, in Lisbon.

If, in other surveys, different models of accelerometers are used, the manufacture company, normally, provides a specific software to analyse the data collected and to export the raw output variables.

In both cases, data analysis should include:
  • Control for normality and homoscedasticity of the study variables;
  • When necessary, log transformations to normalize distributions can be used;
  • Calculate means ± standard deviations;
  • Multiple linear regressions models, adjusted for confounding variables, should be performed to study the independent associations between key variables like age, sex, BMI and other socio-demographic or health markers with relevant physical activity and sedentary behaviour variables;