[W]hen a man breaks down and cries after being asked, “I believe God cares about me and my problems. . . 1 not at all. . . 2 sometimes. . . 3 often. . . 4 always,” it is awfully hard not ask why he is crying.

When I was a graduate student, I collected data for a research study on school success in Mexican American Families. To collect the data, I drove all over Phoenix to meet families in their homes.

I would arrive at the home with two other graduate students. This made a team of three and allowed us to collect data from a mom, a dad, and Jr high student all at the same time. With each of us in a separate room, we would ask 2 hours’ worth of questions from various psychological and sociological measures. We would input the data into a laptop computer. Family members would each receive 60 dollars for their time and data.

This part is important to the blog post: my goal as the survey-giver was to remain neutral. From a scientific viewpoint, my personality, my thoughts, and my influence on the survey takers were all considered “confounds”—messy noise that needed to be eliminated. What the lead researcher really wanted was to find a signal amid the noise. In this study, the signal was any factor that contributed to school success.

Early on in the study, I noticed that noise had a way of asserting itself at every opportunity. At times I wondered if the data we were collecting was valid. It seemed at least possible that a signal could get lost amongst all the noise.

Sometimes, the survey-takers lost focus and answered robotically, like when a student chooses “c” for every question on a multiple-choice test. This was noise. At times people would ask me the meaning of a question—or even how they should answer. This was also noise.

Another thing that happened was people would answer with an answer choice that did not technically exist, like when they would say “Yes” to a question that was supposed to be answered on a scale from “0=not at all” to “5=almost always”. This was surprising to me because we gave the survey-takers a three-ring binder with the answer choices written on sheets of paper. I would prompt them to use the correct paper for each type of question. But, I can’t tell you how many times a person would answer such questions with “Yes” or “No”. Answering “Yes” or “No” instead of one of the 5 answer choices is noise because it meant that I had to interact with the person. The interaction was a deviation from the standard procedure, and my introduction may have introduced more noise.

When people answered “Yes” or “No”, I would repeat the answer choices, or refer the person back to the sheet of paper with the correct choices. My goal was to politely get one of the answer choices without influencing the person. The answer choices were so important because the answer choices corresponded to a numerical value. And once you have a numerical value, you can use statistics to find the signal in the noise.

It was also quite common for a parent to answer a question with advice. So, if the question was, “I listen to my child’s concerns about school work, 0=not at all, 1=sometimes, 2=often, or 3=every day. They would say,

“Oh, it is very important to listen to our children. Sometimes they have a bad day and I can see it in their face, but they don’t want to talk about it, so I leave them alone. . . and then after dinner, while we are cleaning up, I ask a couple of questions . . . “

You guessed it. In the world of hard, factual, materialistic, reductionist science, that whole string of sentences is considered noise. I—and all the other graduate students—were trained to respond in a similar manner to such stories. Because if we each responded in a different way, it would add more noise.

I’ve got to hand it to the lead researchers. They understood that noise happens. And they even developed tables that would tell them how many people they needed to have in a study in order to prevent the noise from canceling out the signal. These tables calculated how loud the signal needed to be in relation to the sample size. A small sample size requires a very loud signal. On the other hand, a large sample size meant that the researchers were willing to accept as valid a weaker signal.

I say all of this as a preface to discussing a tension we are feeling while conducting research into Resilient Recovery Groups.

In our research on Resilient Recovery Groups, our Survey-giver, Anna, is well aware of the importance of scientific objectivity. She attempts to provide the same, confound-free delivery of the questions to each person who takes our surveys. She not only tries to remain neutral and unbiased between different survey-takers, but also between survey times. For example, we give the same survey to people before and after they receive 6 weeks of Resilient Recovery Group. And she tries to maintain the same objective—but helpful—stance that is required of survey-givers for both survey one and survey two.

But when a man breaks down and cries after being asked, “I believe God cares about me and my problems. . . 1 not at all. . . 2 sometimes. . . 3 often. . . 4 always,” it is awfully hard not ask why he is crying.

In this particular case, it turns out they were tears of gratitude and the man eventually responded, “Yes! Yes! I believe he cares!”

To Science, the tears and the explanation are both noise.

Yet because we are Christians, we feel some tension about remaining entirely neutral. In our research study, we believe that every aspect of the research is an opportunity for evangelism. Like post-modernists, we believe it is impossible to remove our bias from our work. Pretending that we are objective is at best silly, and at worst it is lie.

We are beginning to explore the idea of research as evangelism. In one example of this research-as-evangelism paradigm, Anna has found more than once that when asked, “how often have you been to church in the last month,” she has an opportunity to help people find a church. One woman said, “yeah. I went to church on Easter. It’s the first time I’ve been to church since I was 8 years old.” She went on to talk about how she had been angry with God due to traumatic experiences. But it didn’t stop there. She mused out loud about the process of healing—how it takes time, and how she still believes it is possible for her to heal and find peace with God.

Wow.

That’s some noise, there.

But, we’ll take it. We want to know. We want to hear. We believe that simply asking the question and listening might provide an opportunity for further healing and growth in the woman’s spiritual life.

The same question provoked another conversation. This time the woman said, “at this point the only thing stopping me from attending church is a ride.” The consummate scientist would robotically direct the survey-taker back to the answer choices. To do otherwise would risk contaminating the results. If the survey-giver follows up on the nascent desire to attend church, the scientist would never know if the Resilient Recovery Group improved church attendance, or if the conversation with the survey-giver improves church attendance.

In our case, Anna, the Survey-Giver, spent a moment with the woman to help her think of who could provide a ride to church. Why would she do this? We are operating under the controlling idea that it would be unloving and unethical to ignore the woman, and to avoid helping her find an appropriate church.

As we move forward, we will be intentional in our use of surveys. And survey-givers will be instructed to listen to the noise, ask about the noise, and treat the noise as a signal. We will lift up the ignoble noise and treat it with every bit of care we can muster.

After all the God who said, “the first shall be last” and “the greatest among you shall become a servant” cares deeply about the hated, the ignored, the much-maligned. . . noise.