OpenIDEO is an open innovation platform. Join our global community to solve big challenges for social good. Sign up, Login or Learn more

Data for Everyone!

A new tool that offers, for less than $1,000 a year, a level of data management that larger organizations pay thousands of dollars for.

Photo of Alex Kolker
0 2

Written by

Updates: How has your idea changed or evolved throughout the Prize? What updates have you made to this submission? (1500 characters)

Nothing has changed in the design of our innovation through the mentoring process. However, Justin gave us invaluable advice on how to better present the innovations of our model, not only increasing the clarity of this proposal but also more directly expressing how revolutionary the services we have to offer are.

Name or Organization

SuccessLink

Brad McCalla of SuccessLink provides the software design for this data warehouse model. We have already created data warehouses for three different communities, each one designed to suit that community's unique needs. The first prototype, in Waterloo, Iowa, has been in operation for over a decade.

Alex Kolker of Stone Soup Consulting provides guidance on the legal and partnership structures necessary to support such a project.

Geography

We are in Waterloo, IA – but our proposal is to replicate our model in a broad range of communities.

What is your stage of development?

  • Advanced Innovator with 3 to 10+ years of experience in ECD

Type

  • Non-profit

What is the stage of your proposal?

  • Piloting: I have started to implement my solution as a whole with a first set of real users.

Describe how your solution could be a game-changer for your selected Opportunity Area (600 characters)

Most organizations rely on population-level data to identify community need. They collect outcomes data, but rarely have an efficient way to track any peripheral or long-term benefits. Finally, several types of organizations are not allowed to share their data with outside entities.

Our data warehouse model was designed to work around all of these barriers to provide clean, actionable data analyses which: better identify community need, identify likely solutions, and test the short- and long-term impact of those solutions, all for just a few hundred dollars each year.

Select an Innovation Target

  • Service: A new or enhanced service that creates value for end beneficiaries.

Tell us more about your innovation (1500 characters)

Our innovation is not a computer system, commercial service, nor software package. Our innovation is a new approach to data management, consisting of three elements:

a. A data-sharing strategy that makes it possible for outside entities, such as funders and nonprofits, to access data protected by federal regulations such as FERPA and HIPAA.

b. A data-sharing agreement template that communities can alter to suit their unique needs.

c. The linkage of 2 third-party, off-the-shelf software packages that makes this data collection and analysis possible for less than $1,000 per year.

Although many funders and service providers would like better data-management capabilities, they feel such capabilities require (a) the hiring of specialized computer science and/or statistics personnel, (b) the collection of an encyclopedic amount of data, and (c) an investment of thousands of dollars per year. None of these is true.

In its ease of use and low annual cost, our model makes data warehousing practical for nonprofits and funders of any size.

This strategy was first designed for – and is most useful for – funders and providers in the early childhood education sphere.

Our data warehouse model creates a neutral space where FERPA-protected data can be cross-referenced with data from early childhood education programs, so that those programs can track the full academic careers of their past clients, allowing them to collect data on the long-term impact of their services.

What problem are you aiming to solve? (3 sentences)

1. How do you track a program's long-term impacts (for example, grades of children who once participated in a neonatal program)?

2. How do you combine data sets from multiple partners: different types of organizations (preschool programs and school districts, for example) or across boundaries (between multiple school districts, for example)?

3. How do you test which intervention is the most effective for specific sub-populations ("Does this program work better for girls than boys?")?

Explain your idea (5000 characters)

Our data-management strategy makes it possible to collect individual-level student data – not just grades and attendance, but also identifying information such as name and date of birth – for all students in a public school's data system without violating FERPA's confidentiality restrictions.

We can then use our data warehouse to cross-reference this information with data from other entities. For example, a preschool could enter the names of all students who graduated from their program in a certain year. The warehouse could then find those students in school district records and report back an aggregate number: for example, a preschool could find out what percentage of the graduates from their program are reading at grade-level by third grade.

This type of information is useful in three very different ways:

ONE: Identification of need by subpopulation

Our data warehouse model allows end-users to break down their data findings according to any division your data sources can provide. Are male students more successful than female students? Are African-American students more successful than Hispanic students? Are high-income students more successful than low-income students? Are students from one neighborhood more successful than students from another neighborhood?

Being able to divide up the data in so many ways, it is possible to determine which subpopulations are most in need of your services, allowing communities to focus limited resources where they are most needed.

TWO: Creating predictive models

Our data warehouse model makes it possible to identify correlations between performance at different stages of a child's academic career: for example, which early childhood education factor (letter recognition, self-regulation, positive self-image) is the best predictor of which student will be more successful in school (say, reading at grade level by third grade).

Instead of waiting until students fail their third-grade reading proficiency tests to begin to provide services, communities can use these predictive models to identify preschoolers who are at higher risk for reading issues in the future, and provide preventative interventions that keep these students from developing reading problems in the first place. Preventative interventions are not only less expensive, but also more effective.

THREE: Testing interventions

Suppose, as in the example above, we find a correlation between a child's self-regulatory skills at age 3 and a child's ability to read at age 8.

We then take a group of 3-year-olds who have weak self-regulatory skills and divide them into 5 groups.

One group takes part in each of three possible interventions: different approaches to fostering a child's social-emotional growth. The fourth group takes part in a reading program. The fifth group is the control, receiving no intervention at all.

Certainly each program can measure short-term outcomes of their services. But the real payoff won't come until four years later, when those children take their third-grade reading standardized tests. Our data warehouse model makes it easy to track these students as they enter and move up through elementary school.

Immediately, we can answer several questions:

1. Were any of the interventions more effective at producing strong readers than the control?

2. Were the social-emotional programs more effective at producing strong readers than the reading program?

3. Which social-emotional program was the most effective of the three?

But our data warehouse model can do deeper dives than this. For example:

4. Are the children who were most successful at achieving their intervention's short-term goals also more likely to be strong readers in third grade?

5. Are the children who had good attendance during their early childhood intervention more likely to be strong readers than the students with poor attendance?

6. Is a specific intervention more successful with one subpopulation (for example, male versus female students) than another?

All of this information will help communities select the most effective programs, and also to test different ways to improve those programs on an ongoing basis.

You can see how this data warehousing model can take a single question and return so much data that it would take years for the community to act on all of the findings. It is certainly a level of data management that most nonprofits believe is too sophisticated and expensive to be practical. However, we can achieve this high level of functionality for less than $1,000 each year.

Who benefits? (1500 characters)

The most obvious beneficiaries of our data warehouse would be the children served, by agencies able to focus their efforts and resources on those sub-populations who need them most, providing the exact interventions that have proved most effective.

But another beneficiary will be the agencies themselves. Our data warehouse model can provide a level of data-management sophistication that has traditionally been too cost-prohibitive for most nonprofits.

Using off-the-shelf software and online subscription services, we were able to set up our prototype for an up-front cost of approximately $1,960, and will be able to operate the warehouse at an ongoing annual cost of just $630.

These reduced costs make high-level data management and analysis accessible to just about any nonprofit agency in the nation.

A final benefit for the agencies who adopt the model is its usefulness in securing grant funds, both in the writing of grant applications (using up-to-date, local data to demonstrate community need) and the completion of funder outcome reports.

The very first incarnation of the data warehouse model, for example, has earned Waterloo, Iowa, over $5 million dollars in funding over the last six years, including large-scale grants from the Wellmark Blue Zones Fund, Substance Abuse and Mental Health Administration, Health Resources and Services Administration, and 21st Century Learning grants.

What kind of impact will your idea have? (1500 characters)

The foundation of most of our understanding of the educational needs for preschool-aged children comes from two landmark longitudinal studies completed in the 1970's: the Perry Preschool Project and the Abecedarian Early Intervention Project. These studies were expensive and labor-intensive, so they have not been repeated; much of our activity in early childhood education for the past 40 years is still grounded in this early work.

Our data warehouse model not only facilitates longitudinal analysis, but allows you to analyze data for your full population, rather than just for a statistical sample. This eliminates the need for standard deviations and margins of error, creating the most accurate data measurements possible.

In addition, each individual community will be to base their research on the very students they are hoping to serve, allowing them to tailor their interventions to the unique needs of their community.

How does or how could your idea impact low-income children? (1500 characters)

Our data warehouse model is a powerful tool for identifying the impacts of poverty on early childhood development.

At its simplest level, we can track the outcomes for all low-income children as compared to all high-income children in a given population.

But we can also flag cohorts in our system – including picking a random sample to use as a control group. We can then compare the progress of a control group of low-income children to a second cohort of low-income children who received a specific intervention. With enough years of data to draw upon, we can even create these cohorts several years after the fact, allowing for a full-scale longitudinal analysis generated in just a few minutes. This will be a powerful tool in deciding which approaches are most effective in mitigating the needs of a community's low-income population.

More than this, we can dive deeper into the impacts of poverty specific to a unique community. Instead of lumping all low-income children into a single category, we can compare the needs and the effective interventions for one low-income population compared to another. We can, for example, compare the needs and outcomes of low-income children living in two separate neighborhoods. We can compare the needs and outcomes of children who remain in poverty to those whose families move out of poverty later in the child's life. This additional level of precision will provide a more accurate picture of the needs of a community's low-income children.

Innovation: What makes your concept innovative? (5000 characters)

Most early childhood development programs measure the immediate impact of their services – usually through exit surveys or some form of end-of-year assessment. While these data are useful, they cannot prove what the early childhood development programs most want to know: whether or not their services actually impact the long-term development of their clients.

Most providers feel that such longitudinal analysis is (a) impossible, due to federal data restrictions, and (b) prohibitively expensive, requiring an investment in sophisticated software or data-management services.

The innovation of our approach is that it puts high-level data-management capabilities into the hands of the providers, at a cost affordable to even small-scale nonprofit organizations.

Nothing in our model is new by and of itself. Our data-collection strategies use provisions written into FERPA and HIPAA regulations. Our data storage and analysis capabilities come from SQL database management and off-the-shelf data-analysis software.

The true innovation of our model is that we can provide everything in a single, easily-replicated package:
* The research framework to govern a community's collection and analysis of data.
* Data-sharing agreement templates designed to ensure that all data collection and analysis complies with federal confidentiality guidelines.
* The design of a SQL "data universe" to store the data in.
* Methods for standardizing the data collected from different sources.
* Methods for cross-referencing data on the same group of individuals collected from multiple sources.
* The linkage of the SQL database to an information-processing online subscription service.

Scale: Describe how your idea could reach a significant number of end-users. (1500 characters)

This model was designed to reach end-users nationwide. As a powerful yet inexpensive data solution, this model would be accessible to, and of great use to, nearly every community and/or nonprofit organization in the country.

Feasibility: Where are you with understanding the feasibility of your idea? Describe what you’ve done so far and your plans. (3000 characters)

Our data warehouse model was designed for maximum feasibility.

As we noted above, our toolkit for replicating our data warehouse design – research framework, data-sharing agreement templates, software linkages, and data management best practices – comes ready-made: designed to be easily adapted to the needs of any community.

We have already instituted several different iterations of this data-warehousing model in multiple communities across the state of Iowa: Clear Lake, Council Bluffs, Davenport, Mason City, and Waterloo. No two cities have implemented the model in exactly the same way.

There are only two components of our model that the community itself has to provide.

The first component is the partnership structure that determines what data are supplied, who houses them, and what limits there should be on how data analyses from the system can and cannot be used.

The second component is the funds: approximately $1,000, plus travel expenses, for SuccessLink to come set up the data system, and $630-per-year user subscription for the systems data-management software.

We would like to use our OpenIDEO Prize funds to reduce these costs for communities interested in replicating (and testing out) our data-warehousing model. We estimate these costs, on average, at $5,000 per community, meaning that an award of $80,000 would facilitate the creation of data warehouses in 16 communities.

This replication will be mutually beneficial. The communities can defray much of the (already inexpensive) up-front system set-up costs. For our part, we get to test the replicability of our data warehouse model in many different types of communities and applied to conducting research on a wide variety of issues. This will allow us to further fine-tune and increase the effectiveness of our model.

Business Viability: How viable is your business model? (5000 characters)

The main viability of our business model is that both entities – SuccessLink and Stone Soup Consulting – are experienced in the field, having implemented this model in several communities already. Both entities are dedicated to continuing to replicate this model nationwide for the foreseeable future.

In addition, our high-performance-for-low-cost model will be highly competitive in the field. All other standard data management models available – Clear Impact being the current industry leader – cost tens of thousands of dollars more per year to implement, making them cost-prohibitive for the majority of nonprofits in the nation.

HCD: How have you used human centered design to build or refine your concept? (5000 characters)

The adaptability built into our data warehouse design encourages different communities to implement the model in their own unique ways. In other words, each community can adapt our model to the unique situation of the population they are trying to measure.

Each new iteration of the model teaches us more about the versatility of our design and shows us new ways in which the model can be applied to a wide variety of situations.

Tell us more about you (3000 characters)

SuccessLink (http://www.successlink.us) began in 1994 as part of the national Communities In Schools (CIS) network designed to keep youth in school.

The primary objective of CIS was to link local health and human service providers from the Cedar Valley area together, to establish school based health centers known as Success Street. Success Street is a partnership of local providers who bring their service directly into the schools where they can be accessed more easily by students and families.

In order to increase the efficiency of Success Street, SuccessLink oversaw the creation of a community-wide data system.

This system began in 2001 as an attempt to capture data at the individual level in efforts to better track and measure community-wide outcomes. Data is gathered from the area school districts as well as most of the area youth-serving agencies such as Boys & Girls Club, YMCA, YWCA, Big Brothers Big Sisters, etc. The data is gathered and placed into a database where it is then linked together allowing agencies to measure how well their participants are doing in school. Any number of queries can be run once the data is connected, allowing for a much richer and deeper understanding of how local programs are impacting participants. This system has allowed all involved agencies to measure different types of outcomes using a data-driven system.

Alex Kolker of Stone Soup Consulting has adapted SuccessLink's data model to serve as a cradle-to-career database, linking benchmark data for children from birth through post-secondary completion. The goal was to create predictive-analytics profiles that will help communities detect who would benefit from early, preventative services, rather than waiting until the students were already in crisis to provide curative services. The data warehouse model's ability to incorporate social service outcomes data would then indicate which local interventions would be most effective at providing these early preventative services.

Kolker helped design the partnership structure that would allow for the sharing of data and drew up the data-sharing agreements to ensure that the project would not violate federal confidentiality guidelines.

Do you have the people and partners you need to do what you’ve described? (600 characters)

We have the people in place ready to replicate our data warehouse model for any community who wants it.

Alex Kolker will work with communities to build partnerships and negotiate data-sharing agreements to govern data warehouse operations.

Once this is in place, Brad McCalla will personally travel to participating communities to set up their system and train local staff in its use.

Both Kolker and McCalla will be available to help trouble-shoot the system after it is launched. They will maintain a network of participating communities to share best practices and suggest improvements.

As you consider your next steps, what kinds of help could you use? Is there a type of expertise that would be most helpful? (1800 characters)

There are several ways that experts could be of use to us.

Although our template data-sharing agreements have been vetted by local attorneys, we would still appreciate a review of these documents by experts in data management law.

Our software structure is using off-the-shelf software for purposes for which it was not originally designed. It would be interesting if people with programming expertise could design similar software specifically for the purpose of implementing our data warehouse model.

Finally, we would welcome any suggestions for improving our system design from experts in statistical analysis.

Even if we do not win an OpenIDEO prize, input from such experts during the mentoring phase of the Challenge would be most welcome.

Would you like mentoring support?

  • Yes

If so, what type of mentoring support do you think you need? (1200 characters)

There are several ways that mentors could be of use to us.

Although our template data-sharing agreements have been vetted by local attorneys, we would still appreciate a review of these documents by experts in data management law.

Our software structure is using off-the-shelf software for purposes for which it was not originally designed. It would be interesting if people with programming expertise could design similar software specifically for the purpose of implementing our data warehouse model.

Finally, we would welcome any suggestions for improving our system design from experts in statistical analysis.

Are you willing to share your email contact information submitted on OpenIDEO with Gary Community Investments?

  • Yes, share my contact information

[Optional] Biography: Upload your biography. Please include links to relevant information (portfolio, LinkedIn profile, organization website, etc).

Brad McCalla has been the Director of SuccessLink since June, 1995. He graduated from Hamline University in St. Paul, MN with a degree in Psychology and Sociology. He also holds an MS and Ed. S in Higher Education Administration from Drake University.

Alex Kolker has a Ph.D. in English Language and Literature from the University of Kansas. He has been working in the non-profit sphere for over a decade, specializing in collective impact strategies and the design of social-service initiatives.

Mentorship: How was your idea supported? (5000 characters)

Our mentor was Justin Liang, Financial Technology Executive at LendingClub. We had a single phone call that was casual and chatty in nature: three data geeks all excited about a neat new idea.

Justin was able to point us towards a couple of organizations who are doing similar work – most notably the Ed-Fi Alliance (https://www.ed-fi.org). Justin was able to confirm, however, that our idea was not only viable but also definitely a new approach.

The best help that Justin gave us, though, was in his role as a knowledgeable but objective observer: guiding us into better, more clear, and more direct ways to explain what exactly is innovative in our data warehouse model. We believe that this new approach in the current version of the proposal will be not only more understandable, but also raise higher levels of enthusiasm, to the non-technology-inclined professionals who read it.

Updates: How has your idea changed or evolved throughout the Prize? What updates have you made to this submission? (1500 characters)

Nothing has changed in the design of our innovation through the mentoring process. However, Justin gave us invaluable advice on how to better present the innovations of our model, not only increasing the clarity of this proposal but also more directly expressing how revolutionary the services we have to offer are.

Name or Organization

SuccessLink

Brad McCalla of SuccessLink provides the software design for this data warehouse model. We have already created data warehouses for three different communities, each one designed to suit that community's unique needs. The first prototype, in Waterloo, Iowa, has been in operation for over a decade.

Alex Kolker of Stone Soup Consulting provides guidance on the legal and partnership structures necessary to support such a project.

Geography

We are in Waterloo, IA – but our proposal is to replicate our model in a broad range of communities.

What is your stage of development?

  • Advanced Innovator with 3 to 10+ years of experience in ECD

Type

  • Non - Profit

What is the stage of your proposal?

  • Piloting: I have started to implement my solution as a whole with a first set of real users.

Describe your submission in one clear sentence

Our data warehousing model is a powerful, versatile tool that offers, for less than a thousand dollars a year, a level of data management that larger organizations pay tens of thousands of dollars to achieve.

Describe how your solution could be a game-changer for your selected Opportunity Area (600 characters)

Most organizations rely on population-level data to identify community need. They collect outcomes data, but rarely have an efficient way to track any peripheral or long-term benefits. Finally, several types of organizations are not allowed to share their data with outside entities.

Our data warehouse model was designed to work around all of these barriers to provide clean, actionable data analyses which: better identify community need, identify likely solutions, and test the short- and long-term impact of those solutions, all for just a few hundred dollars each year.

Select an Innovation Target

  • Service: A new or enhanced service that creates value for end beneficiaries.

Tell us more about your innovation (1500 characters)

Our innovation is not a computer system, commercial service, nor software package. Our innovation is a new approach to data management, consisting of three elements:

a. A data-sharing strategy that makes it possible for outside entities, such as funders and nonprofits, to access data protected by federal regulations such as FERPA and HIPAA.

b. A data-sharing agreement template that communities can alter to suit their unique needs.

c. The linkage of 2 third-party, off-the-shelf software packages that makes this data collection and analysis possible for less than $1,000 per year.

Although many funders and service providers would like better data-management capabilities, they feel such capabilities require (a) the hiring of specialized computer science and/or statistics personnel, (b) the collection of an encyclopedic amount of data, and (c) an investment of thousands of dollars per year. None of these is true.

In its ease of use and low annual cost, our model makes data warehousing practical for nonprofits and funders of any size.

This strategy was first designed for – and is most useful for – funders and providers in the early childhood education sphere.

Our data warehouse model creates a neutral space where FERPA-protected data can be cross-referenced with data from early childhood education programs, so that those programs can track the full academic careers of their past clients, allowing them to collect data on the long-term impact of their services.

What problem are you aiming to solve? (3 sentences)

1. How do you track a program's long-term impacts (for example, grades of children who once participated in a neonatal program)?

2. How do you combine data sets from multiple partners: different types of organizations (preschool programs and school districts, for example) or across boundaries (between multiple school districts, for example)?

3. How do you test which intervention is not only the most effective for which specific sub-populations ("Does this program work better for girls than

Explain your idea (5000 characters)

Our data-management strategy makes it possible to collect individual-level student data – not just grades and attendance, but also identifying information such as name and date of birth – for all students from a public school's data system without violating FERPA's confidentiality restrictions.

We can then use our data warehouse to cross-reference this information with data from other entities. For example, a preschool could enter the names of all students who graduated from their program in a certain year. The warehouse could then find those students in school district records and report back an aggregate number: for example, a preschool could find out what percentage of the graduates from their program are reading at grade-level by third grade.

This type of information is useful in three very different ways:

ONE: Identification of need by subpopulation

Our data warehouse model allows end-users to break down their data findings according to any division your data sources can provide. Are male students more successful than female students? Are African-American students more successful than Hispanic students? Are high-income students more successful than low-income students? Are students from one neighborhood more successful than students from another neighborhood?

Being able to divide up the data in so many ways, it is possible to determine which subpopulations are most in need of your services, allowing communities to focus limited resources where they are most needed.

TWO: Creating predictive models

Our data warehouse model makes it possible to identify correlations between performance at different stages of a child's academic career: for example, which early childhood education factor (letter recognition, self-regulation, positive self-image) is the best predictor of which student will be more successful in school (say, reading at grade level by third grade).

Instead of waiting until students fail their third-grade reading proficiency tests to begin to provide services, communities can use these predictive models to identify preschoolers who are at higher risk for reading issues in the future, and provide preventative interventions that keep these students from developing reading problems in the first place. Preventative interventions are not only less expensive, but also more effective.

THREE: Testing interventions

Suppose, as in the example above, we find a correlation between a child's self-regulatory skills at age 3 and a child's ability to read at age 8.

We then take a group of 3-year-olds who have weak self-regulatory skills and divide them into 5 groups.

One group takes part in each of three possible interventions: different approaches to fostering a child's social-emotional growth. The fourth group takes part in a reading program. The fifth group is the control, receiving no intervention at all.

Certainly each program can measure short-term outcomes of their services. But the real payoff won't come until four years later, when those children take their third-grade reading standardized tests. Our data warehouse model makes it easy to track these students as they enter and move up through elementary school.

Immediately, we can answer several questions:

1. Were any of the interventions more effective at producing strong readers than the control?

2. Were the social-emotional programs more effective at producing strong readers than the reading program?

3. Which social-emotional program was the most effective of the three?

But our data warehouse model can do deeper dives than this. For example:

4. Are the children who were most successful at achieving their intervention's short-term goals also more likely to be strong readers in third grade?

5. Are the children who had good attendance during their early childhood intervention more likely to be strong readers than the students with poor attendance?

6. Is a specific intervention more successful with one subpopulation (for example, male versus female students) than another?

All of this information will help communities select the most effective programs, and also to test different ways to improve those programs on an ongoing basis.

You can see how this data warehousing model can take a single question and return so much data that it would take years for the community to act on all of the findings. It is certainly a level of data management that most nonprofits believe is too sophisticated and expensive to be practical. However, we can achieve this high level of functionality for less than $1,000 each year.

Who benefits? (1500 characters)

The most obvious beneficiaries of our data warehouse would be the children served, by agencies able to focus their efforts and resources on those sub-populations who need them most, providing the exact interventions that have proved most effective.

But another beneficiary will be the agencies themselves. Our data warehouse model can provide a level of data-management sophistication that has traditionally been too cost-prohibitive for most nonprofits.

Using off-the-shelf software and online subscription services, we were able to set up our prototype for an up-front cost of approximately $1,960, and will be able to operate the warehouse at an ongoing annual cost of just $630.

These reduced costs make high-level data management and analysis accessible to just about any nonprofit agency in the nation.

A final benefit for the agencies who adopt the model is its usefulness in securing grant funds, both in the writing of grant applications (using up-to-date, local data to demonstrate community need) and the completion of funder outcome reports.

The very first incarnation of the data warehouse model, for example, has earned Waterloo, Iowa, over $5 million dollars in funding over the last six years, including large-scale grants from the Wellmark Blue Zones Fund, Substance Abuse and Mental Health Administration, Health Resources and Services Administration, and 21st Century Learning grants.

What kind of impact will your idea have? (1500 characters)

The foundation of most of our understanding of the educational needs for preschool-aged children comes from two landmark longitudinal studies completed in the 1970's: the Perry Preschool Project and the Abecedarian Early Intervention Project. These studies were expensive and labor-intensive, so they have not been repeated; much of our activity in early childhood education for the past 40 years is still grounded in this early work.

Our data warehouse model not only facilitates longitudinal analysis, but allows you to analyze data for your full population, rather than just for a statistical sample. This eliminates the need for standard deviations and margins of error, creating the most accurate data measurements possible.

In addition, each individual community will be to base their research on the very students they are hoping to serve, allowing them to tailor their interventions to the unique needs of their community.

How does or how could your idea impact low-income children? (1500 characters)

Our data warehouse model is a powerful tool for identifying the impacts of poverty on early childhood development. At its simplest level, we can track the outcomes for all low-income children as compared to all high-income children in a given population.

But we can also flag cohorts in our system – including picking a random sample to use as a control group. We can then compare the progress of a control group of low-income children to a second cohort of low-income children who received a specific intervention. With enough years of data to draw upon, we can even create these cohorts several years after the fact, allowing for a full-scale longitudinal analysis generated in just a few minutes. This will be a powerful tool in deciding which approaches are most effective in mitigating the needs of a community's low-income population.

More than this, we can dive deeper into the impacts of poverty specific to a unique community. Instead of lumping all low-income children into a single category, we can compare the needs and the effective interventions for one low-income population compared to another. We can, for example, compare the outcomes of low-income children living in two separate neighborhoods. We can compare the outcomes of children who remain in poverty to those whose families move out of poverty later in the child's life. This additional level of precision will provide a more accurate picture of the needs of a community's low-income children.

Innovation: What makes your concept innovative? (1500 characters)

Most early childhood development programs measure the immediate impact of their services – usually through exit surveys or some form of end-of-year assessment. While these data are useful, they cannot prove what the early childhood development programs most want to know: whether or not their services actually impact the long-term development of their clients.

Most providers feel that such longitudinal analysis is (a) impossible, due to federal data restrictions, and (b) prohibitively expensive, requiring an investment in sophisticated software or data-management services.

The innovation of our approach is that it puts high-level data-management capabilities into the hands of the providers, at a cost affordable to even small-scale nonprofit organizations.

The true innovation of our model is that we can provide everything in a single, easily-replicated package:
* The research framework to govern a community's collection and analysis of data.
* Data-sharing agreement templates designed to ensure that all data collection and analysis complies with federal confidentiality guidelines.
* The design of a SQL "data universe" to store the data in.
* Methods for standardizing the data collected from different sources.
* Methods for cross-referencing data on the same group of individuals collected from multiple sources.
* The linkage of the SQL database to an information-processing online subscription service.

Scale: Describe how your idea could reach a significant number of end-users. (1500 characters)

This model was designed to reach end-users nationwide. As a powerful yet inexpensive data solution, this model would be accessible to, and of great use to, nearly every community and/or nonprofit organization in the country.

Feasibility: Where are you with understanding the feasibility of your idea? Describe what you’ve done so far and your plans. (3000 characters)

Our data warehouse model was designed for maximum feasibility.

As we noted above, our toolkit for replicating our data warehouse design – research framework, data-sharing agreement templates, software linkages, and data management best practices – comes ready-made: designed to be easily adapted to the needs of any community.

We have already instituted several different iterations of this data-warehousing model in multiple communities across the state of Iowa: Clear Lake, Council Bluffs, Davenport, Mason City, and Waterloo. No two cities have implemented the model in exactly the same way.

There are only two components of our model that the community itself has to provide.

The first component is the partnership structure that determines what data are supplied, who houses them, and what limits there should be on how data analyses from the system can and cannot be used.

The second component is the funds: approximately $1,000, plus travel expenses, for SuccessLink to come set up the data system, and $630-per-year user subscription for the systems data-management software.

We would like to use our OpenIDEO Prize funds to reduce these costs for communities interested in replicating (and testing out) our data-warehousing model. We estimate these costs, on average, at $5,000 per community, meaning that an award of $80,000 would facilitate the creation of data warehouses in 16 communities.

This replication will be mutually beneficial. The communities can defray much of the (already inexpensive) up-front system set-up costs. For our part, we get to test the replicability of our data warehouse model in many different types of communities and applied to conducting research on a wide variety of issues. This will allow us to further fine-tune and increase the effectiveness of our model.

Business Viability: How viable is your business model? (1500 characters)

The main viability of our business model is that both entities – SuccessLink and Stone Soup Consulting – are experienced in the field, having implemented this model in several communities already. Both entities are dedicated to continuing to replicate this model nationwide for the foreseeable future.

In addition, our high-performance-for-low-cost model will be highly competitive in the field. All other standard data management models available – Clear Impact being the current industry leader – cost tens of thousands of dollars more per year to implement, making them cost-prohibitive for the majority of nonprofits in the nation.

HCD: How have you used human centered design to build or refine your concept? (1500 characters)

The adaptability built into our data warehouse design encourages different communities to implement the model in their own unique ways. In other words, each community can adapt our model to the unique situation of the population they are trying to measure.

Each new iteration of the model teaches us more about the versatility of our design and shows us new ways in which the model can be applied to a wide variety of situations.

Tell us more about you (3000 characters)

SuccessLink (http://www.successlink.us) began in 1994 as part of the national Communities In Schools (CIS) network designed to keep youth in school.

The primary objective of CIS was to link local health and human service providers from the Cedar Valley area together, to establish school based health centers known as Success Street. Success Street is a partnership of local providers who bring their service directly into the schools where they can be accessed more easily by students and families.

In order to increase the efficiency of Success Street, SuccessLink oversaw the creation of a community-wide data system.

This system began in 2001 as an attempt to capture data at the individual level in efforts to better track and measure community-wide outcomes. Data is gathered from the area school districts as well as most of the area youth-serving agencies such as Boys & Girls Club, YMCA, YWCA, Big Brothers Big Sisters, etc. The data is gathered and placed into a database where it is then linked together allowing agencies to measure how well their participants are doing in school. Any number of queries can be run once the data is connected allowing for a much richer and deeper understanding of how local programs are impacting participants. This system has allowed all involved agencies to measure different types of outcomes using a data-driven system.

Alex Kolker of Stone Soup Consulting has adapted SuccessLink's data model to serve as a cradle-to-career database, linking benchmark data for children from birth through post-secondary completion. The goal was to create predictive-analytics profiles that will help communities detect who would benefit from early, preventative services, rather than waiting until the students were already in crisis to provide curative services. The data warehouse model's ability to incorporate social service outcomes data would then indicate which local interventions would be most effective at providing these early preventative services.

Kolker helped design the partnership structure that would allow for the sharing of data and drew up the data-sharing agreements to ensure that the project would not violate federal confidentiality guidelines.

Do you have the people and partners you need to do what you’ve described? (500 characters)

We have the people in place ready to replicate our data warehouse model for any community who wants it.

Alex Kolker will work with communities to build partnerships and negotiate data-sharing agreements to govern data warehouse operations.

Once this is in place, Brad McCalla will personally travel to participating communities to set up their system and train local staff in its use.

Both Kolker and McCalla will be available to help trouble-shoot the system after it is launched.

As you consider your next steps, what kinds of help could you use? Is there a type of expertise that would be most helpful? (1800 characters)

There are several ways that experts could be of use to us.

Although our template data-sharing agreements have been vetted by local attorneys, we would still appreciate a review of these documents by experts in data management law.

Our software structure is using off-the-shelf software for purposes for which it was not originally designed. It would be interesting if people with programming expertise could design similar software specifically for the purposes of our model.

Finally, we would welcome any suggestions for improving our system design from experts in statistical analysis.

Even if we do not win an OpenIDEO prize, input from such experts during the mentoring phase of the Challenge would be most welcome.

Would you like mentoring support? [Relevant only for Early Submission Deadline]

  • Yes

If so, what type of mentoring support do you think you need? (1200 characters) [Relevant only for Early Submission Deadline]

There are several ways that mentors could be of use to us.

Although our template data-sharing agreements have been vetted by local attorneys, we would still appreciate a review of these documents by experts in data management law.

Our software structure is using off-the-shelf software for purposes for which it was not originally designed. It would be interesting if people with programming expertise could design similar software specifically for the purposes of our model.

Finally, we would welcome any suggestions for improving our system design from experts in statistical analysis.

Even if we do not win an OpenIDEO prize, input from such experts during the mentoring phase of the Challenge would be most welcome.

Mentorship: How was your idea supported? [Relevant only for our early submission participants] (1500 characters)

Our mentor was Justin Liang, Financial Technology Executive at LendingClub. We had a single phone call that was casual and chatty in nature: three data geeks all excited about a neat new idea.

Justin was able to point us towards a couple of organizations who are doing similar work – most notably the Ed-Fi Alliance (https://www.ed-fi.org). Justin was able to confirm, however, that our idea was not only viable but also definitely a new approach.

The best help that Justin gave us, though, was in his role as a knowledgeable but objective observer: guiding us into better, more clear, and more direct ways to explain what exactly is innovative in our data warehouse model. We believe that this new approach in the current version of the proposal will be not only more understandable, but also raise higher levels of enthusiasm, to the non-technology-inclined professionals who read it.

Are you willing to share your email contact information submitted on OpenIDEO with Gary Community Investments?

  • Yes, share my contact information

[Optional] Biography: Upload your biography. Please include links to relevant information (portfolio, LinkedIn profile, organization website, etc).

Brad McCalla has been the Director of SuccessLink since June, 1995. He graduated from Hamline University in St. Paul, MN with a degree in Psychology and Sociology. He also holds an MS and Ed. S in Higher Education Administration from Drake University.

Alex Kolker has a Ph.D. in English Language and Literature from the University of Kansas. He has been working in the non-profit sphere for over a decade, specializing in collective impact strategies and the design of social-service initiatives.

Attachments (1)

SuccessLink Graphic.pdf

A graphic of our data warehouse design.

0 comments

Join the conversation:

Comment