Project Context

Overview

Discover & Research

While many young (e.g. under thirty) adult Americans understand the importance of their credit score, few take a hands-on on approach to managing it. A 2014 study conducted by Student Monitor [1] reflected this claim, in which the large majority (74%) of college students in the participant sample did not know their personal credit scores. This finding indicated that most students also do not take a proactive approach to managing their scores. Understanding how a credit score can affect future opportunities may be confusing for young adults who are newly independent. Young financial independents are not consistently aware of what affects their score[1].

In this project, we asked, how can we help educate this audience on the importance of managing their credit scores? Specifically, what types of technology-based products might enable better understanding and monitoring of credit scores for newly financially independent people; additionally, we asked, what kinds of features would be important to include? 

In the United States, a credit score is comprised of personal information from three different scores. The critical output is known as the FICO Score, which is a calculated overview that summarizes an objective overview of fiscal trustworthiness [2]. “Contributors” refer to these separate scores that are a part of the FICO credit score. “Credit sources” are the applications and companies that offer people insight to these scores and contributors. Today, consumer credit cards are adding complimentary credit score reports to their service to help their customers understand the important factors about credit scores. 

Discover Card, an established credit institution, enables card members to view their FICO score at no additional cost on their phones [7]. While made easily accessible to card members, the information provided is limited to one contributor. Additionally, these reports focus solely on current scores and only include the period that the cardholder has been a member. 

There are also independent, online ways to get full views of personal financial health. Mint.com is a free web-based product that aims to create more transparency around spending habits [4].

Research Methods

This study had three sequential components: an observational study, an interview study, and a survey study. Conclusions from each study informed the direction and emphasis of the next, building a more detailed picture of the problem space and opportunities for design recommendations. 

Observational Study

Participants

We recruited four participants through our individual social networks; two were 23-year-old women and two were men (26 and 28 years old). All participants were college educated. Three participants had maintained financial independence for less than two years as of January 2016, and one had been financially independent for more than 3 years. See Table 1 for participant details.

Data collection

We first debriefed out participants about the research topic and obtained consent to proceed (see Appendix for consent form). We then prompted participant to interpret and act upon: “Could you show me how you would go about finding your credit score?” For this round of inquiry, our goal was to observe in a fly-on-the-wall manner: observing actions and taking notes using the AEIOU framework, with minimal influence on participant behavior. We attended to their activities (A), environment (E), interactions with others (I), and the objects they used (O). 

Data analysis

Observational notes were entered onto virtual sticky notes in the online StormBoard tool. We created an affinity diagram to organize points of observations sequentially. Sequences were grouped based on similarity and co-occurrence among participants.

Interview Study 

In the following sections, we describe our interview participants and our data collection and analysis methods. 

Participants

We recruited four participants through our social and work networks; two men, aged 23 and 26, and two women ages 24 and 25. All participants were college educated and had maintained financial independence for less than four years. See Table 2 for participant details.

Data collection

We began our interviews by reading our interview script to each participant. The participant answers and paraphrased their responses. 

In the interviews, we explored the usage of checking credit scores by asking:

• If the participants were currently financially independent or not.

• How often participants check their credit scores, if at all.

• Participant knowledge on calculations of the credit score and their understanding of the credit system.

• The amount of time each participant had been financially independent.

• The way the participants perceive the effect of the credit system towards their life.

Data analysis

Team members organized participant responses in a spreadsheet, corresponding visually with interview questions. Interviewer observations and comments were alongside responses, which provided further clarity to the context and “gut feelings” the interviewers had during the interviews. After the spreadsheets had completed interview data, we transposed responses into Mural, an online whiteboard-collaboration tool. With the data in one place, we began to surface themes and behavior patterns. We clustered responses, thematically and formulated our findings, which would inform our survey.

Survey  

The observation and interview studies provided the team with a glimpse of user behaviors in this problem space. In the following sections, we describe our methods for the final study conducted: the survey study. Behaviors and patterns from former studies framed the questionnaire used in this survey.

Participants

Given the themes and patterns we identified in earlier studies, the sample size for this study needed enough power to be statistically significant, which we decided would be best as a sample of over 30 people. We recruited through the DePaul University Participant Pool and our immediate social and professional networks. After leaving the survey active for 5 days, we received 83 responses.

Data collection

In the survey, we explored the validity of our themes from earlier studies by asking:

• How long has the participant been financially independent?

• In terms of their financial well-being, what are participants concerned about?

• Has the participant ever checked their credit score? If yes, how often in the last year?

• What factors does the participant think affect their credit scores?

• Has the participant ever been denied from a credit card or loan? If yes, did they conduct any research about the credit system to better the odds of being approved in the future?

Data analysis

With an exported document of raw survey data from Google Forms, we first translated the data into a form that would work for SPSS calculations. 

• Changing text strings to numerical data (Answers from “Which of the following do you believe affect your personal credit score? Check all that apply”) and tabulating responses into ordinal data. Otherwise known in this paper as “level of understanding about which factors affect credit,” (UFAC). Value represented a score, with a possible maximum score: 13.

• For the question, “Have you ever been denied a credit?” “Never applied for a loan” got changed to “No”

• Participants who claimed they financially independent for less than three years of got grouped into a single group (original ranges: 0 to 6 months, 6 months to 1 year, 1 to 2 years)

Asking “Which of the following do you believe affect your personal credit score? Check all that apply” targeted our dependent variable for our first hypothesis. Asking if the participant has been denied a from a credit card or loan aimed to define our independent variable for our second hypothesis.

• The previous tests used : Mann-Whitney U were not capable to display data thoroughly or to prove the hypothesis as true or false, used a different test (t-test, but Levene test to compare populations or groups) 

• Years spent as financially independent need to be recalculated because of the skew presented in the participants

• Participants that chose “I have never applied” were added into group 2 (those were not denied) 

• Levene’s test: tested the assumption that there was equal of variance in the sample. 

Findings (Approach Research)

Observational Study

Our synthesis of observations led us to believe that there were multitudes of effective ways to not only acquire credit scores, but also become aware of what affects the scores. Additionally, we found there was a steep learning curve in understanding of how to get credit scores among our participants. The FICO score was the most commonly referenced score. Participants who had sought out their scores previously acted quickly and acquired their score without hesitation. Participants who were experienced in locating their credit score(s) had various reason for choosing their platform.

• Most (3 out of 4) people used smart phone as means of finding credit information.

• All participants utilized different platforms and applications.

• One participant expressed having trouble locating their credit score(s), but was eventually able to locate it; they chose a tool through trial and error.

• All participants eventually attained credit scores using either Mint.com or personal credit card mobile applications.

During the observations, participants reflected about their methods to finding credit. Discourse about the credit source tools, which handle sensitive personal information, focused on opinions of trustworthiness. The credibility of credit sources relied heavily on third party reviews about security and general brand-recognition.

Interview Study

Centralizing responses across individuals helped elevate high-level themes. We translated themes into continuous scales, or spectrums; we described the “low” and “high” attributes for each theme/spectrum. The table below describes our themes/spectrums:

We then aligned interviewee responses along the spectrums. In doing this, we began to identify commonalities between participants based on their responses. When commonalities were salient enough, we established patterns between participants and grouped behavioral traits, accordingly. These patterns informed the personas we developed.

The two major user groups from the interview study were (1) people who accurately knew details about the credit systems and (2) people who were unfamiliar with the credit system. Table 3 outlines the differences that we saw between the two groups.

We set forth with the assumptions that there were two distinctly separate groups, which helped collate our hypotheses and informed our survey study:

• Individual’s understanding of credit scores increases as time increase with financial independence.

• There is a correlation between being denied from a credit card or loan and knowledge of the individual's credit score.
   

Survey 

The body of survey responses lacked significantly equal amounts of variance, meaning there were enough inconsistencies in the data to invalidate the use of differential statistics. This was found through conducting Levene’s Tests and determining the variance using the significance outputs. In the following sections, we explain our findings under both hypotheses.

Hypothesis 1: The amount of time spent as financially independent has a positive relationship with the level of understanding about which factors affect credit (UFAC) (via scores from survey question).

Table 4.1 provides descriptive statistics about the two groups that we compared. As shown, the group with less than three years of financial independent (Group A) scored an average UFAC score of 6.7 points out of a possible 13 points, while the group with more than three years of independency (Group B) had an average score of 7.2. This illustrated that the groups had a similar overall UFAC scores, which would disprove our hypothesis.

 The Levene's test (Table 4.2) indicated that we could not test the population samples equally, therefore showing that we could not assume a relationship between UFAC scores and time being financially independent (F = 0.442). The average UFAC scores of individuals that were financially independent (M = 6.8) was less than the group of individuals that were financially independent for more than 3 years (M =7.2). The differences were not significant amongst the two independent groups in UFAC scores: t(16.595) = -.788, p > 0.05.

Hypothesis 2: Having been denied from a credit card or loan correlates to a better understanding about which factors affecting credit (via UFAC scores).

The group statistics (Table 5.1)  for the second set of independent variables (those who were denied a loan with those that have not been denied a loan) showed little variance as well, based on the means and standard deviation. 

The Levene’s test (Table 5.2) indicated no significant difference (p > 0.05) between the two groups, therefore we rejected the null hypothesis and could compare both populations to analyze their knowledge of the credit score system. Simply put: both groups got similar UFAC scores, regardless of whether they were denied after applying for a credit card or loan, or not. The average UFAC score (M = 7.0, SD=2.2) of participants that were denied a credit score is a little less than those that were not denied a credit card or loan (M = 7.1, SD=1.8). Consequently to the lack of significance, the entire t-test was not a valid measure for this hypothesis.

This study found that participants who were denied a credit card or loan had an equal understanding of the factors that influence credit scores (via UFAC score) with those that were not denied a credit card or loan. Given these results, we concluded that no evidence supported that the amount of time spent as financially independent determines the level of understanding of what influences credit scores.

Reflections (Conclusion)

From the findings and identified limitations of the study outlined in this paper, we are able to create a list of implications for design and recommendations for future or continuing work. These implications aim to create a more consistent way of understanding financial information.

Future Work

Were this project to continue or be reconstructed, we recommend a more precise approach to recruitment to help get a balanced sample: one that is proportionally balanced between people who are newly independent (Less than two years) and those who have been independent for more than two years. Furthermore, carefully rewording the definition of “financial independence” may allow participants to categorize themselves more effectively. Conducting research on these same hypotheses may yield similar results, however, only a study with a usable/accurate sample will be able to determine that.

We also recommend other detail about best practices, with the aim that this “future” study group has a more robust understanding of the problem space and user groups:

• “Periodic diary study,” otherwise known as Experience sampling. Ask participant users to record their emotions and assumptions before, during, or after interacting with their credit score in any form. This will inform design about emotional engagement and aim to increase use adoption.

• Source from larger, more diverse set of participants through a peer-reviewed screener in the beginning of each study phase.

Explore around user workarounds, strategies, and self-education that has been effective for our more informed participants.