Sound data management is important not only to ensure quality analysis but also to protect the confidentiality of participants. Typically, communities conducting PiT Counts put data management protocols in place prior to implementing their counts. This ensures that all physical and electronic data is organized and protected. A Youth Count is no different.

The storage of the raw data is an important matter that should be handled carefully. Online documents should be locked and password protected. Protocols should be in place that indicate who has access to data, including survey forms, and for what purpose.

You should agree ahead of time on who owns the data gathered from this process and who has access to it. Access can be granted partially to certain components of the data as well. The lead organization may have full control of the data, while other partners can sign data sharing agreements. – Excerpt from the COH PiT Count Toolkit

A Focus on Youth

For some, data input and analysis can be daunting. However, Youth Count data can be easily managed and interpreted with a good foundational knowledge of Excel and an understanding of basic mathematical techniques – such as means, medians and cross tabulation (many of which can be researched online). Nonetheless, we encourage you to utilize the resources in your community. Partner with students, academics or researchers in the social sciences, preferably ones familiar with survey design and implementation.    

Before the Count

Though data input and analysis happens after the Youth Count, you will need to develop a data management plan well in advance of the count. As you develop your protocol, consider the following: 

TABLE 11. DEVELOPING A DATA MANAGEMENT PLAN

Security of Survey Forms	Survey forms, including photocopies, scanned versions and incomplete forms must be kept securely. How will you secure the completed forms during and immediately after the count? Where will survey forms be kept in the days and weeks following the count? When will they be destroyed? How will they be destroyed? Who will have access to them?
Security of Electronic Data	Once the surveys are inputted, you must take steps to secure all raw data. Carefully consider who will have access to the data. Where will it be stored? How will it be shared?
Ownership	Issues of data ownership should be resolved well before data collection takes place. The Youth Count Committee should agree, in writing, to who owns the data and how access will be granted. The owner of the data should be responsible for maintaining all security measures agreed upon by the Committee.
Data Input	How many completed surveys do you expect to get? Who will input the data? The Youth Count Coordinator? Youth Count partners? Volunteers? Who will oversee the process?
Data Input Template	We recommend that you create a data input template in Excel. If you plan to analyse the data in SPSS or another statistical software package, ensure that the template can be easily imported. Note: most PiT Count analysis can be done in Excel.
Data Input Protocol	If more than one person is inputting data, we recommend that you create a data input protocol. A set of instructions that explains how to resolve common issues such as: determining the eligibility of survey participants, ambiguous responses and conflicting information.
Analysis	Who will conduct the data analysis? Use your Youth Count survey to develop a number of research questions in advance of the analysis. Consult your Youth Count Committee and partners to determine what information will be useful.
Deduplication	As noted previously, there is a greater chance of duplication in a Youth Count than there is in a PiT Count. We recommend that you use unique identifiers, which will facilitate deduplication during data analysis.
Volunteer Training	Once you have developed your data management plan, consider what role your volunteers will play in maintaining the security of data and ensuring high quality data collection. Effective training for all volunteers – whether they are surveying, helping out at headquarters or inputting data – will decrease the likelihood of errors.

After the Count

Once the count is complete, take the following steps:

1. Review the survey forms and tally sheets

As soon after the count as possible, ideally the day after, go through the paper surveys and check for missing entries, ambiguous answers and anomalies. Contact volunteers, while the information is still fresh, to resolve any issues.

2. Input data

It is not necessary to catch all errors or inconsistences prior to the data input stage. There will be additional opportunities to clean the data. Start inputting the data, based on the protocol you established prior to the count. If volunteers are inputting the data, ensure they receive sufficient training beforehand. The Youth Count Coordinator must be available to answer questions and resolve issues while data input is taking place.

3. Keep a record of decisions

Write down everything! Keep a record of decisions that are made during data input. For example: if a youth selects more than one gender, how did you input the response? A detailed record will give you something to refer to, to ensure consistency. This is especially important if more than one person is inputting data.

Helpful Resource After the Count: Data Entry and Cleaning (PPT) –Christina Maes Nino, Community Animator, Social Planning Council of Winnipeg *Available on the COH Workspace on Homelessness

4. Determine eligibility

It is likely that ineligible participants will complete the survey. During the input and analysis phase, remove anyone that does not fit your criteria. For instance, a youth who indicates that they are have a permanent residence and are staying with their parents on the night of the count should be removed from the dataset. Youth over the age of 24 should also be removed.

5. Remove duplicates

Remove surveys with unique identifiers that appear more than once. Compare other data points to remove likely duplicates that were not caught by the unique ID system. If two entries seem suspiciously similar, compare the survey times and locations. Only remove entries if you are certain they are redundant.

Helpful Resource After the Count: Data Entry and Cleaning (PPT) –Christina Maes Nino, Community Animator, Social Planning Council of Winnipeg *Available on the COH Workspace on Homelessness

6. Data cleaning

Check your dataset for illogical values or conflicting information. For example, if a 16-year-old youth indicates that they first experienced homelessness at age 18, you know this is an error. You should refer back to the paper survey to determine whether this is a recording error or a data input mistake. Remove entries that cannot be corrected. Remember to record this decision.

7. Analyzing Data

Once the dataset is complete, you can analyze the data. Basic analysis can be done in Excel. You may require statistical software to do more advanced analysis. Refer to the research questions you established at the outset of the count. Which types of information will be most useful in understanding youth homelessness in your community? Refer to PiT Count reports for useful examples.

Building Alignment: Tips & Strategies

At the outset of planning, you will need to determine whether the Youth Count data will be analysed separately from the general PiT Count results. Will the datasets be combined as one? The general PiT Count should report all individuals experiencing homelessness, including youth. However, there is merit in conducting a complimentary analysis with the surveys collected from youth participants. Without this information, it will not be possible to create a discrete Youth Count report.

Note: if the general PiT Count and the Youth Count use different screening criteria, there may be discrepancies in the reported number of youth experiencing homelessness. For instance, the general PiT Count organizers may choose not to report on hidden homelessness, whereas Youth Count organizers are strongly encouraged to include forms of hidden homelessness, such as couch surfing. To minimize discrepancies, we encourage you to work with general PiT Count organizers to develop consistent screening criteria.