Building Bridges and Bonds - Just Beginning and DadTime Impact Evaluation

The Just Beginning Intervention is comprised of 5 sessions, which each include the following three program components.

(1) Parent Learning Session. A trained and certified facilitator introduces the father to a fundamental but simple concept that may help him to improve the quality of his interactions with his young child. The father watches a short collection of Sesame Beginnings video clips featuring Baby Elmo and other characters that illustrate these concepts and lay the groundwork for the father-child play session. The father and facilitator discuss the video and plan for the father-child play session.

(2) Father-Child Play Session. The father joins his young child in a child-friendly play space, which is safe for the child to explore and contains a colorful floormat, books, toys, and other items that facilitate father-child interactions. The father has an opportunity to try out the new approaches he has just learned as he plays with his child. The facilitator simply observes the interaction and does not interfere with it or judge the father.

(3) Debrief Session. The facilitator describes positive examples of how the father used any new approaches with his child. The father and facilitator reflect on the play session together. The father and facilitator jointly brainstorm ideas for how the father will apply the lesson he learned to other settings.

DadTime uses mobile technology to provide planning tools to support attendance at Just Beginning sessions, and reinforcements of Just Beginning program content.

(1) Program Attendance Planning Tools. DadTime sends an automated reminder one day before a scheduled Just Beginning session. It includes the time and location of the session and gives the father an opportunity to plan for his transportation and travel time. An option allows the father to send an automated text message reminder to the person accompanying the child to the session, if he is not bringing the child himself. Another option allows the father to instantly notify program staff if he needs to reschedule.

(2) Content Support. DadTime sends the father exercises and activity suggestions after each Just Beginning session, to help him reflect on what he has learned and plan for time he may spend with his child between sessions. Several weeks after enrolling in the program, the father begins to receive biweekly prompts that encourage him to plan for continued playtime with his child, as well as reinforce key program concepts.

2016-10-11

2019-03-31

*** For the Just Beginning impact evaluation we have 10 primary outcomes ***:
(1) Frequency that father uses labeling with child.
(2) Father attitude toward verbal engagement with child.
(3) Warm and supportive parenting scale
(4) Father-child dysfunctional interaction from the Parenting Stress Index
(5) Parenting efficacy
(6) Overall quality of father-child relationship
(7) Father-child relationship quality scale
(8) Father perceived influence on child
(9) Father/child contact
(10) Father's reliability for planned time with child

*** For the DadTime impact evaluation we have 3 primary outcomes ***:
(1) Completed first Just Beginning session
(2) Number of Just Beginning sessions completed
(3) Completed all Just Beginning sessions

Notes on the construction of the primary outcome variables appear below.

Primary outcome measures for the Just Beginning impact evaluation:
(1) Frequency that father uses labeling with child.
- Created from the survey question, "In past month, how often did you talk to child about the things that they looked at, grabbed, or pointed to?" which includes categories: (1) more than once a day, (2) about once a day, (3) a few times a week, (4) a few times a month, (5) rarely, and (6) not at all in the past month.
- Survey respondents are not asked this question if they have not seen their child at all in the past month. They are coded as category (6) not at all in the past month.

(2) Father attitude toward verbal engagement with child.
- Created from the survey question, "Do you agree or disagree with this statement: there is not much point talking to my child, because he or she is too young to understand me" which includes response categories: (1) strongly agree, (2) agree, (3) disagree, and (4) strongly disagree. The outcome is created by dichotomizing these responses into: (0) strongly agree or agree and (1) strongly disagree or disagree.

(3) Warm and supportive parenting scale (an average of 4 items).
- Created from the survey questions (1) "In the past 30 days, how often did you hug or show physical affection to your child?"; (2) "In the past 30 days, how often did you praise your child?"; (3) "In the past 30 days, how often did you soothe your child if he or she was crying?"; (4) "In the past 30 days, how often did you tell your child you loved him or her?" These questions all have response categories of (1) more than once a day, (2) about once a day, (3) a few times a week, (4) a few times a month, (5) rarely, and (6) not at all in the past month.
- Survey respondents are not asked this question if they have not seen their child at all in the past 30 day. They are coded as category (6) not at all in the past month, for all questions.
- Before the average is taken to compute the final outcome, responses to each item are reverse-coded, so responses range from (1) not at all in the past month to (6) more than once a day.

(4) Father-child dysfunctional interaction, a subscale from the Parenting Stress Index (average of 11 items).
- Created from 11 survey questions with responses of (1) strongly agree, (2) agree, (3) not sure, (4) disagree, and (5) strongly disagree. The 11 items are not included here because of potential copyright restrictions.
- Before the average is taken to compute the final outcome, responses to each item are reverse-coded, so responses range from (1) strongly disagree to (5) strongly agree.

(5) Parenting efficacy (average of 5 items)
- Created from the average of responses to five survey questions: (1) "I am good at helping my child when he/she is upset or distressed"; (2) "I am good at knowing what activity my child enjoys"; (3) "I am good at getting my child to have fun with me"; (4) "I am good at getting my child to understand what I want him/her to do"; and (5) "I am good at understanding what my child wants or needs." For each question, respondents indicate how often they agree with the statement, with responses of (1) always or almost always, (2) often, (3) sometimes, (4) rarely, and (5) never.
- Before the average is taken to compute the final outcome, responses to each item are reverse-coded, so responses range from (1) never to (5) always or almost always.

(6) Overall quality of father-child relationship.
- Created from response to survey question "Do you feel that your relationship with your child is excellent, very good, somewhat good, fair, or poor?" Respondents can answer (1) excellent, (2) very good, (3) somewhat good, (4) fair, or (5) poor.
- Responses are reverse-coded and categories 4 and 5 are combined for the final outcome, so that the outcome has values (1) poor or fair to (4) excellent.

(7) Father-child relationship quality scale (average of 6 items).
- Created from responses to survey questions (1) "How often do you feel disappointed with your child?"; (2) "How often do you wish that your child was different?"; (3) "How often do you feel proud of your child?"; (4) "How often do you feel angry or irritated with your child?"; (5) "How often do you accept your child the way he/she is?"; and (6) "How often does being a father to your child bring you joy?" Responses to each question are (1) always or almost always, (2) often, (3) sometimes, (4) rarely, and (5) never.
- Before computing the average of the responses to the six questions for the overall outcome, question (3), (5), and (6) are reverse-coded.

(8) Father perceived influence on child (average of 2 items).
- Created from responses to survey questions (1) "How much influence do you think you have on your child's life right now?" and (2) "How much influence do you think you will have on your child's life over the long-term?" Responses to each question are (1) a great deal, (2) some, (3) a little, and (4) none.
- The final outcome is created by averaging the responses to both questions.

(9) Father/child contact
- Created from response to the survey question, "In the past 30 days, how often did you see your child in person?" which has responses of (1) every day or almost every day, (2) 3 or 4 times per week, (3) 1 or 2 times per week, (4) 2 or 3 times in the past month, (5) once in the past month, and (6) not at all.
- Fathers who live with their child all or most of the time are not asked this question, and are coded as (1) every day or almost every day. Fathers who have not seen their child in the past month (asked in a previous question) are not asked this question, and they are coded as if they responded (6) not at all to this question.
- Categories (4) and (5) are combined for the final outcome, such that it has five categories: (1) every day or almost every day, (2) 3 or 4 times per week, (3) 1 or 2 times per week, (4) 1 to 3 times in the past month, and (5) not at all.

(10) Father's reliability for planned time with child
- Created from response to the survey question, "Sometimes things come up that get in the way of plans to spend time with children. How often do you have to cancel plans with child?" which includes responses of (1) often, (2) sometimes, (3) rarely, and (4) never.

For the DadTime impact evaluation we have 3 primary outcomes:
(1) Completed first Just Beginning session
- Binary variable equal to 1 if the father has completed the first Just Beginnings session and 0 otherwise.

(2) Number of Just Beginning sessions completed
- A count of the number of Just Beginning sessions the father has completed (range from 0 to 5).

(3) Completed all Just Beginning sessions
- Binary variable equal to 1 if the father has completed all 5 Just Beginning sessions and 0 otherwise.

For the Just Beginning impact evaluation we have 35 secondary outcomes:
(1) Father-child dysfunctional interaction - single item self-rating for good parenting
(2) Father-child engagement in caregiving activities
(3) Father-child engagement in learning and play activities
(4) Father-child engagement in physical play activities
(5) Father-child engagement in socialization activities
(6) Harsh discipline
(7) Physical harsh discipline
(8) Parenting efficacy scale: financial items
(9) Parenting efficacy rating: Following through on promises
(10) Decision-making influence
(11) Father dedication/commitment to child scale
(12) Father report of how much he would miss his child
(13) Father-child relationship quality scale: positive items
(14) Father-child relationship quality scale: negative items
(15) Current perceived influence rating
(16) Future perceived influence rating
(17) Father-child contact not in person rating
(18) Nights in the past month father-child spent in same house
(19) Focal child lives with father
(20) Father lives with any of his children
(21) Contact with any children in past 30 days
(22) Cooperative co-parenting rating: "Mother and I make a good parenting team"
(23) Cooperative co-parenting rating: "Mother tells me I am doing a good job"
(24) Co-parenting challenges: undermining scale
(25) Co-parenting challenges: conflict scale
(26) Co-parenting challenges: maternal gate-keeping scale
(27) Perceived stress scale: helplessness
(28) Perceived stress scale: efficacy
(29) Total amount of child support paid in the past month
(30) Child support owed versus paid in the past month
(31) Number of types of informal support provided to focal child in the past month
(32) Number of types of informal support provided to non-focal children in the past month
(33) Positive discipline
(34) Currently employed
(35) Amount worked since random assignment

For the DadTime impact evaluation we have 4 secondary outcomes:
(1) Father/child contact
(2) Father's reliability for planned time with child
(3) Frequency that father uses labeling with child
(4) Warm and supportive parenting scale

Notes on the construction of the secondary outcome variables appear below.

Secondary outcome measures for the Just Beginning impact evaluation:
(1) Father-child dysfunctional interaction - a single item self-rating of good parenting
- Measure is created by taking the father's rating on a five-item scale where he is asked to rate whether he is (1) a very good parent to (5) not very good at being a parent.

(2) Father-child engagement in caregiving activities
- Measure created as the average of responses to 8 questions asking how often the father has engaged in certain caregiving activities with his focal child over the past 30 days.

(3) Father-child engagement in learning and play activities
- Measure created as the average of responses to 4 questions asking how often the father has engaged in particular learning and play activities with his focal child over the past 30 days.

(4) Father-child engagement in physical play activities.
- Measure created as the average of responses to 7 questions asking how often the father has engaged in particular learning and play activities with his focal child over the past 30 days.

(5) Father-child engagement in socialization activities
- Measure created as the average of 4 questions asking how often the father has engaged in particular socialization activities with his focal child over the past 30 days.

(6) Harsh discipline
- Measure created by averaging frequency of 4 types of harsh discipline the father has used in the past three months with his focal child.

(7) Physical harsh discipline
- Measure created by averaging the frequency of two types of physical harsh discipline the father has used in the past three months with his focal child.

(8) Parenting efficacy scale: financial items
- Taken as the average of father's ratings of how frequently he agrees with two statements: (1) "I am good at providing for [my child's] financial needs" and (2) "I am good at providing diapers, milk, or other needs for [my child]."

(9) Parenting efficacy rating: following through on promises
- Taken as the father's rating of how frequently he agrees with the statement: "I am good at following through with my promises to [my child]."

(10) Decision-making influence
- Measure is the father's rating of how much influence he feels he has over major decisions for his focal child's life. He can respond: (1) no influence, (2) some influence, or (3) a great deal of influence.

(11) Father dedication/commitment to child scale
- Measure is the average of father's responses of how much he agrees with each of five statements about how important his children and being a father are to him. Fathers can provide responses to each question of (1) strongly agree, (2) agree, (3) disagree, or (4) strongly disagree.

(12) Father report of how much he would miss his child
- Father rating of how much he would miss his focal child if he could not see them in the next month, with values: (1) a great deal and (2) less than a great deal.

(13) Father-child relationship quality scale: positive items
- Created from responses to survey questions (1) "How often do you feel proud of your child?"; (2) "How often do you accept your child the way he/she is?"; and (3) "How often does being a father to your child bring you joy?" Responses to each question are (1) always or almost always, (2) often, (3) sometimes, (4) rarely, and (5) never.
- Before computing the average of the responses to the 3 questions for the overall outcome, question responses are reverse-coded.

(14) Father-child relationship quality scale: negative items
- Created from responses to survey questions (1) "How often do you feel disappointed with your child?"; (2) "How often do you wish that your child was different?"; and (3) "How often do you feel angry or irritated with your child?” Responses to each question are (1) always or almost always, (2) often, (3) sometimes, (4) rarely, and (5) never.
- Final measure is taken as the average of responses to all 3 questions.

(15) Current perceived influence rating
- Father rates how much influence he feels he has on his focal child's life right now, ranging from (1) a great deal to (4) none.

(16) Future perceived influence rating
- Father rates how much influence he feels he will have on his focal child's in the future, ranging from (1) a great deal to (4) none.

(17) Father-child contact not in person rating
- Father responds how often he has communicated with his focal child not in-person. Fathers can respond from (1) every day or almost every day to (5) not at all.
- Fathers responding "my child is too young for this" are treated as if they responded (5) not at all. Fathers who live with their focal child all or most of the time have their own category (6) for this outcome.

(18) Nights in the past month father-child spent in same house
- Father indicates how many nights he and his focal child have spent together in the past 30 days, ranging from 0 to 30.

(19) Focal child lives with father
- Binary measure of whether the father lives with his focal child all or most of the time.

(20) Father lives with any of his children
- Binary measure of whether the father lives with any of his children all or most of the time.

(21) Contact with any children in past 30 days
- Number of children father has seen in person in the last 30 days, with categories (1) saw all of his children, (2) saw some of his children, or (3) saw none of his children.

(22) Cooperative co-parenting rating: " Co-parent and I make a good parenting team"
- Binary measure of whether father agrees or disagrees with the above statement about the focal child.

(23) Cooperative co-parenting rating: "Co-parent tells me I am doing a good job"
- Binary measure of whether father agrees or disagrees with the above statement about the focal child.

(24) Co-parenting challenges: undermining scale
- Father rates how strongly he agrees with four statements about the co-parent undermining his parenting activities for his focal child.
- Outcome is created as the average of responses across all items.

(25) Co-parenting challenges: conflict scale
- Father rates how strongly he agrees with four statements about his getting into conflicts or arguments about parenting with the co-parent of his focal child.
- Outcome is created as the average of responses across all items.

(26) Co-parenting challenges: maternal gate-keeping scale
- Father rates how strongly he agrees with two statements about the co-parent's getting in the way of his engagement with his focal child.
- Outcome is created as the average of responses across all items.

(27) Perceived stress scale: helplessness
- Father rates how often in the past month he has felt helpless, which is asked in a series of six questions or scenarios. Responses range from (1) never to (5) very often.
- Outcome is created as the average of responses across all six items.

(28) Perceived stress scale: efficacy
- Father rates how often in the past month he has felt effective, which is asked in a series of four questions or scenarios. Responses range from (1) never to (5) very often.
- Outcome is created as the average of responses across all four items.

(29) Total amount of child support paid in the past month
- Continuous measure of the amount of formal child support the father actually paid in the past month. Fathers who are not required to pay any child support are treated as if they paid $0.

(30) Child support owed versus paid in the past month
- Categorical measure of the amount of formal child support the father paid compared to the amount he owed. Categories are (1) not required to pay any amount, (2) paid none of what was required, (3) paid some of what was required, (4) and paid all of or more than what was required.

(31) Number of types of informal support provided to focal child in the past month
- Fathers indicate whether they provided seven different types of informal support for their focal child. Final measure can range from 0 to 7.
- Fathers who do not live with their focal child all or most of the time are missing on this outcome, and so will not be included in its analysis.

(32) Number of types of informal support provided to non-focal children in the past month
- Fathers indicate whether they provided seven different types of informal support for their non-focal children. Final measure can range from 0 to 7.
- Fathers with only one child or who do not live with any children all or most of the time are missing on this outcome, and so will not be included in its analysis.

(33) Positive discipline
- Fathers rate how often in the past 3 months they have practice 3 types of positive discipline, ranging from (0) never to (5) more than 10 times. Final measure is the average of responses across the three questions.
- Fathers who have not seen their focal child in the past 3 months are coded as (0) never.

(34) Currently employed
- Binary outcome equal to whether the father currently has a job (including full- or part-time jobs or temporary, transitional, or seasonal jobs).

(35) Amount worked since random assignment
- Categorical outcome of how much respondent has worked since random assignment, ranging from (1) all the time to (6) never.

For the DadTime impact evaluation we have 4 secondary outcomes, which are all primary outcomes for the Just Beginnings impact evaluation. They are created as indicated in the "primary outcomes (explanation)" section above.

The Just Beginning parenting intervention was added to the services offered at three existing fatherhood programs. Fathers who sought services at one of these programs were screened for eligibility for the Just Beginning program. Fathers who reported having recent contact with a child of theirs who was between 2 months and under three years of age was eligible for the Just Beginning program component, unless there was a legal restriction or other extenuating circumstance that would prohibit their participation with their child. Eligible fathers were randomly assigned to a Just Beginning program group, which is offered the new Just Beginning program component in addition to the usual services at the fatherhood program, or to a "Services as Usual" control group, which does not have access to the new Just Beginning program component.

In the same three sites testing the Just Beginning parenting intervention, the Building Bridges and Bonds study also tested a new smartphone app, called DadTime, to encourage fathers’ program engagement. The DadTime app was installed on smartphones of fathers in the Just Beginning program group upon enrollment.

For the DadTime engagement test, fathers who are in the Just Beginning program group went through a second stage of random assignment to a “Full DadTime” app group or a “Partial DadTime” app group. The Full DadTime group was offered access to the DadTime app content from the time of intake onward, and could use the app both during and after their period of participation in Just Beginning sessions. The Full DadTime group received both of the two key DadTime components: 1) reminders and attendance support, including pop-up reminders about JB session times, and short questions and suggestions to help fathers plan attendance at the sessions; and 2) activity suggestions, including short questions and suggestions to help plan when they will see their children and activities they can do together; the app tailored questions and suggestions based on the child’s age and how frequently the father said he saw his child.

In contrast, the Partial DadTime group did not have access to DadTime during their period of Just Beginning participation, and did not receive reminders or attendance support. The Partial DadTime group did, however, receive the same activity suggestions as the Full DadTime group starting about eight weeks after enrolling in the Building Bridges and Bonds study, which was intended to be after the father had completed his five Just Beginning sessions.

After eight weeks, the DadTime app sent suggestions to fathers of ways to play with their children, which were designed to potentially extend all Just Beginning fathers’ practice of the Just Beginning parenting skills. The study team selected the eight-week interval to set a common timeframe when both the Full and Partial DadTime groups would likely have completed their Just Beginning sessions. Fathers in the Services as Usual control group were not be eligible for the app because the app is designed to reinforce skills learned in Just Beginning.

Randomization was done by computer.

Individuals

No

Fathers were be randomly assigned within 1 of 3 program sites.

738 fathers are in the Just Beginning study. 224 fathers are in the DadTime study.

For the Just Beginning study, 368 fathers were randomly assigned to the Just Beginning program group and 370 fathers were randomly assigned to the Services as Usual control group.

For the DadTime study, 117 fathers were randomly assigned to the Full DadTime group and 107 fathers were randomly assigned to the Partial DadTime group.

For the Just Beginning impact study, the impact analysis will use follow-up survey data. We expect to have follow-up survey data for 80 percent of fathers in the study. The minimum detectable effect size estimate is .177 standard deviations. For a binary outcome variable, we are powered to detect impacts of around 9 percentage points. For the DadTime study, we will rely on administrative records data from the MIS that tracks program participation. Therefore, we expect to have outcome data for all 224 fathers. We are powered to detect impacts that are at least .289 standard deviations. We are unlikely to find impacts of that magnitude. Therefore, the DadTime impact study is considered exploratory.