Participants will reflect on anticipated benefits of purchasing healthy foods, consistent with motivational interviewing and self-determination theory. Message content will be personalized as follows: During the initial workshop, all participants will complete an exercise identifying benefits of healthy eating that are important to them. Message content will then be programmed to be personalized according to anticipated rewards important to that participant. Participants assigned to have this intervention component OFF will not have content about anticipated benefits of change added to any messages.

Coaches will monitor participant food purchases via a dashboard for viewing purchase data, and send messages designed to provide feedback and enhance supportive accountability for program goals. The messages will provide reinforcement for purchases consistent with program goals and express concern for areas in which adherence is low. Participants assigned to have this component OFF will not receive these extra messages or phone calls, and their food purchases will only be viewed by research staff for research outcome assessment purposes.

If participants are assigned to have household support ON, one adult in the household will receive weekly text messages designed to elicit support for changing food purchases. In addition, the index participant and household member will be invited to participate in one extra workshop session and three brief coaching calls focused on household support. If participants are assigned to have this OFF, household members will have no program involvement.

In Wave 1, the primary outcome will be change from 0-20 weeks in dietary intake of the major food categories from the American Institute for Cancer Research guidelines. Self-reported dietary recalls will be completed for three days at baseline and week 20 using the Automated Self Administered 24-hour food recall (ASA-24). This table reports the change in ounces of processed meat consumed per week.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

In Wave 1, the primary outcome will be change from 0-20 weeks in dietary intake of the major food categories from the American Institute for Cancer Research guidelines. Self-reported dietary recalls will be completed for three days at baseline and week 20 using the Automated Self Administered 24-hour food recall (ASA-24). This table reports the change in grams of added sugars consumed per day.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

In Wave 1, the primary outcome will be change from 0-20 weeks in dietary intake of the major food categories from the American Institute for Cancer Research guidelines. Self-reported dietary recalls will be completed for three days at baseline and week 20 using the Automated Self Administered 24-hour food recall (ASA-24). This table reports the change in mg of sodium consumed per day.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

In Wave 1, the primary outcome will be change from 0-20 weeks in dietary intake of the major food categories from the American Institute for Cancer Research guidelines. Self-reported dietary recalls will be completed for three days at baseline and week 20 using the Automated Self Administered 24-hour food recall (ASA-24). This table reports the change in grams of saturated fat consumed per day.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

In Wave 1, the primary outcome will be change from 0-20 weeks in dietary intake of the major food categories from the American Institute for Cancer Research guidelines. Self-reported dietary recalls will be completed for three days at baseline and week 20 using the Automated Self Administered 24-hour food recall (ASA-24). This table reports the change in oz of sugar sweetened beverages consumed per day.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

In Wave 2, the Diet History Questionnaire (DHQ-III) will be used instead of food recalls to assess dietary intake. This table reports the change in grams of fiber per day.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

In Wave 2, the Diet History Questionnaire (DHQ-III) will be used instead of food recalls to assess dietary intake. This table reports the change in cups of fruits and vegetables per day.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

In Wave 2, the Diet History Questionnaire (DHQ-III) will be used instead of food recalls to assess dietary intake. This table reports the change in oz of red meat consumed per week.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

In Wave 2, the Diet History Questionnaire (DHQ-III) will be used instead of food recalls to assess dietary intake. This table reports the change in oz of processed meat consumed per week.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

In Wave 2, the Diet History Questionnaire (DHQ-III) will be used instead of food recalls to assess dietary intake. This table reports the change in grams of added sugars consumed per day.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

In Wave 2, the Diet History Questionnaire (DHQ-III) will be used instead of food recalls to assess dietary intake. This table reports the change in mg of sodium consumed per day.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

In Wave 2, the Diet History Questionnaire (DHQ-III) will be used instead of food recalls to assess dietary intake. This table reports the change in grams of saturated fat per day.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

In Wave 2, the Diet History Questionnaire (DHQ-III) will be used instead of food recalls to assess dietary intake. This table reports the change in ounces of sugar-sweetened beverages per day.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

Supportive accountability was measured using 5 items adapted from the Perceptions of Accountability Scale (Chhabria et al., 2020). Participants rated 5 items (e.g., "I feel accountable to others for meeting my dietary goals") on a sliding scale from 0 (not at all) to 100 (very much). Scores on the 5 items were averaged to produce a global Supportive Accountability total score (range 0-100), with higher scores indicating higher perceived supportive accountability to others for dietary/healthy eating goals.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

Motivation for healthy eating was measured using 10 items adapted from the Treatment Self-Regulation Questionnaire (Levesque et al., 2007). Using the question stem "The reason I would follow a healthy diet is...", participants rated items on a scale from 1 (not at all true) to 7 (very true). Six items were averaged to create a subscale of autonomous Motivation for healthy eating (range 1-7), with higher scores on each subscale indicated higher Autonomous Motivation.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

Motivation for healthy eating was measured using 10 items adapted from the Treatment Self-Regulation Questionnaire (Levesque et al., 2007). Using the question stem "The reason I would follow a healthy diet is...", participants rated items on a scale from 1 (not at all true) to 7 (very true). Four items were averaged to create the subscale of external Motivation for healthy eating (range 1-7). Higher scores on each subscale indicated higher external motivation for healthy eating.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

Participants responded to 10 items adapted from the Sallis Social Support for Diet Questionnaire (Sallis et al., 1987). Using the question stem "The other adult(s) in my household...", participants rated items on a scale from 1 (none) to 5 (very often), with 6 indicating "Not Applicable" (N/A; recoded as 1). Five items were summed to measure encouragement of healthy eating (e.g., "complimented me on changing my eating habits"; range 5-25), with higher scores indicating higher encouragement. This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

Participants completed the Sallis Social Support for Diet Questionnaire (Sallis et al., 1987). Using the question stem "The other adult(s) in my household...", participants rated items on a scale from 1 (none) to 5 (very often). 6 indicated "Not Applicable," which was recoded as 1 (none). Five items (e.g., "ate unhealthy foods in front of me") were summed to create the Discouragement of Healthy Eating subscale (range 5-25); higher scores indicated higher discouragement.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

Quality of the household relationship was measured using 5 items adapted from the Relationship Assessment Scale (Hendrick, 1988). Items assessed the quality of their relationship with the person who served as their household support partner in the Eatwell program (e.g., "How well does this person meet your needs") on a scale from 1 (Low) to 5 (High). Scores on the 5 items were averaged to calculate a Relationship Quality total score (range 1-5); higher scores indicated higher perceived relationship quality.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

Weight history was measured using an investigator developed questionnaire. Participants answered the following question about weight history in the past 6 months: "Which of these statements best describe what has happened to your weight during the past 6 months?" Answer choices included: (1) My weight has stayed the same; (2) I've been losing weight; (3) I've been gaining weight; or (4) My weigh has fluctuated a lot.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

Dietary restraint was measured using the Three Factor Eating Questionnaire (Cappelleri et al., 2009). For items #1-20, participants rated statements on a 4-point Likert scale from 1 (definitely true) to 4 (definitely false; reverse coded). Item 21 assessed overall restraint on an 8-point scale from 1 (no restraint) to 8 (total restraint), recoded such that 1-2=1; 3-4=2; 5-6=3; and 7-8=4 (so that all items ranged 1-4). Six items were summed to calculate a Restraint score (range = 6-24); higher scores indicated more restraint.

This pilot study used a factorial design, with 16 conditions and ~4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

Uncontrolled eating was measured using the subscale of the Three Factor Eating Questionnaire - 21 item (Cappelleri et al., 2009). Responses ranged from 1 (definitely true) to 4 (definitely false). Items were reverse coded and summed to calculate an uncontrolled eating (range = 6-24), with higher scores indicating higher uncontrolled eating.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

Emotional eating was measured using the subscale of the Three Factor Eating Questionnaire - 21 item (Cappelleri et al., 2009). Responses ranged from 1 (definitely true) to 4 (definitely false). Six were reverse coded and summed to calculate an emotional eating (range = 6-24), with higher scores indicating higher emotional eating.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

A 13-item Food Frequency Questionnaire was adapted from similar measures (Rifas-Shiman et al., 2001). To measure intake for guideline 1 (i.e., fruits, vegetables, whole grains, and bean/ legumes ), participants rated how frequently they consumed these categories (4 items) during an average week over the past month (0=Never, 1=less than once per week, 2=once per week, 3=2-4 times/week, 4=nearly daily/daily, 5=greater than twice per day). A total guideline 1 score was calculated as the average of responses to the 4 items (range 0-5), where higher scores indicate more frequently eating fruits, vegetables, whole grains, and beans/legumes. This study used a factorial design, with 16 conditions (with ~4 participants) and was designed to collapse across conditions (i.e., comparing 8 conditions where factor is ON to 8 OFF conditions). Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

A 13-item Food Frequency Questionnaire was adapted from similar measures (Rifas-Shiman et al., 2001). To measure intake for guideline 2 (i.e., processed foods high in fat, starches, sugars), participants rated how frequently they consumed foods from these categories (6 items) during an average week over the past month (0=Never, 1=less than once per week, 2=once per week, 3=2-4 times/week, 4=nearly daily/daily, 5=greater than twice per day). A total guideline 2 score was calculated as the average of responses to the 6 items (range 0-5), where higher scores indicate more frequently eating processed foods. This study used a factorial design, with 16 conditions (with ~4 participants) and was designed to collapse across conditions (i.e., comparing 8 conditions where factor is ON to 8 OFF conditions). Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

A 13-item Food Frequency Questionnaire was adapted from similar measures (Rifas-Shiman et al., 2001). To measure intake for guideline 3 (i.e., red and processed meat), participants rated how frequently they consumed these categories (2 items) during an average week over the past month (0=Never, 1=less than once per week, 2=once per week, 3=2-4 times/week, 4=nearly daily/daily, 5=greater than twice per day). A total guideline 3 score was calculated as the average of responses to the 2 items (range 0-5), where higher scores indicate more frequently eating red and processed meats. This study used a factorial design, with 16 conditions (with ~4 participants) and was designed to collapse across conditions (i.e., comparing 8 conditions where factor is ON to 8 OFF conditions). Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

A 13-item Food Frequency Questionnaire was adapted from similar measures (Rifas-Shiman et al., 2001). To measure intake for guideline 4 (i.e., sugar-sweetened drinks), participants rated how frequently they consumed these drinks (1 item) during an average week over the past month (0=Never, 1=less than once per week, 2=once per week, 3=2-4 times/week, 4=nearly daily/daily, 5=greater than twice per day). Given intake for guideline 4 was assessed with only 1 item on the FFQ, the total guideline 4 score was the response to this single item (range 0-5), where higher scores indicate more frequently drinking sugar-sweetened drinks. This study used a factorial design, with 16 conditions (with ~4 participants) and was designed to collapse across conditions (i.e., comparing 8 conditions where factor is ON to 8 OFF conditions). Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

An adapted 8- item version of the Treatment Acceptability Questionnaire (Hunsley, 1993) was used to assess intervention acceptability. Eight questions probed their program experience and how helpful/acceptable they found various intervention components (e.g., "How effective do you think this program is?"). Responses ranged from 1 = no/low acceptability to 7 = high acceptability. A total score was calculated as the average across all 8 items (range 8-56) where higher scores indicate greater treatment acceptability.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

An adapted 8- item version of the Treatment Acceptability Questionnaire (Hunsley, 1993) was used to assess intervention acceptability. Eight questions probed their program experience and how helpful/acceptable they found various intervention components (e.g., "How effective do you think this program is?"). Responses ranged from 1 = no/low acceptability to 7 = high acceptability. A total score was calculated as the average across all 8 items (range 8-56) where higher scores indicate greater treatment acceptability.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

Two self-report items developed by the investigator assessed the goals of the selected household member at baseline: "On a scale of 1-100, how important is it that you adopt healthier eating habits right now?". Item 2 asked "On a scale of 1-100, how important is it that you help your household member adopt healthier eating habits right now?". For each item, household members responded on a scale from 0 - not important at all, 50 - about as important as most of the other things I would like to achieve now, to 100 - most important thing in my life now. Items were analyzed at individually, and each item ranged 0-100 where higher scores indicate higher importance of the associated goal.

Household members completed the same 13-item Food Frequency Questionnaire as the index participant (adapted from Rifas-Shiman et al., 2001) to measure dietary intake. Household members are presented with 13 categories of foods that are relevant for each of the cancer prevention dietary guidelines (e.g., "Fresh, frozen, or canned fruit (not dried or fruit juice)", "Red meats such as beef, pork, lamb"). They rated how frequently they consume each category during an average week over the past month. Scale is as follows: 0 = Never, 1 = less than per week, 2= once per week, 3 = 2-4 times per week, 4 = nearly daily or daily, 5 = greater than twice per day. The 13 items are then classified into the guideline they pertain to and total guidelines scores were calculated as the average of responses to all relevant items (range 0-5) where higher scores indicate more frequently eating items associated with that nutrition guideline.

Household members completed the same adapted 8- item version of the Treatment Acceptability Questionnaire (Hunsley, 1993) as index participants. This measured acceptability of the intervention. Household members were presented with 8 questions about their experience during the program and how helpful/acceptable they found various intervention components to be (e.g., "How effective do you think this program is?", "How trustworthy do you think your coach was?". Household members responded to each question on a 7-point Likert scale where 1 = no/low acceptability and 7 = high acceptability. A total score was calculated as the average across all 8 items (range 8-56) where higher scores indicate greater treatment acceptability.

Household members completed the same 13-item Food Frequency Questionnaire as the index participant (adapted from Rifas-Shiman et al., 2001) to measure dietary intake. To measure intake consistent with guideline 1 of the dietary cancer prevention guidelines (i.e., intake of fruits, vegetables, whole grains, and bean/ legumes ), household members rated how frequently they consumed these categories (4 items) during an average week over the past month. Scale was as follows: 0 = Never, 1 = less than once per week, 2= once per week, 3 = 2-4 times per week, 4 = nearly daily or daily, 5 = greater than twice per day. A total guideline 1 score was calculated as the average of responses to the 4 items (range 0-5), where higher scores indicate more frequently eating fruits, vegetables, whole grains, and beans/legumes.

Household members completed the same 13-item Food Frequency Questionnaire as the index participant (adapted from Rifas-Shiman et al., 2001) to measure dietary intake. To measure intake consistent with guideline 2 of the dietary cancer prevention guidelines (i.e., intake of processed foods high in fat, starches, or sugars), household members rated how frequently they consumed foods that fell into these categories (6 items) during an average week over the past month. Scale was as follows: 0 = Never, 1 = less than once per week, 2= once per week, 3 = 2-4 times per week, 4 = nearly daily or daily, 5 = greater than twice per day. A total guideline 2 score was calculated as the average of responses to the 6 items (range 0-5), where higher scores indicate more frequently eating processed foods high in far, starches, and sugars.

Household members completed the same 13-item Food Frequency Questionnaire as the index participant (adapted from Rifas-Shiman et al., 2001) to measure dietary intake. To measure intake consistent with guideline 3 of the dietary cancer prevention guidelines (i.e., intake of red and processed meats), household members rated how frequently they consumed red meats and processed meats (2 items) during an average week over the past month. Scale was as follows: 0 = Never, 1 = less than once per week, 2= once per week, 3 = 2-4 times per week, 4 = nearly daily or daily, 5 = greater than twice per day. A total guideline 3 score was calculated as the average of responses to the 2 items (range 0-5), where higher scores indicate more frequently eating red and processed meats.

Household members completed the same 13-item Food Frequency Questionnaire as the index participant (adapted from Rifas-Shiman et al., 2001) to measure dietary intake. To measure intake consistent with guideline 4 of the dietary cancer prevention guidelines (i.e., intake of sugar sweetened beverages), household members rated how frequently they consumed sugar sweetened beverages (1 item) during an average week over the past month. Scale was as follows: 0 = Never, 1 = less than once per week, 2= once per week, 3 = 2-4 times per week, 4 = nearly daily or daily, 5 = greater than twice per day. A total guideline 4 score was calculated as the response to this item (range 0-5), where higher scores indicate more frequently drinking sugar sweetened beverages.

To calculate adherence to the NCI dietary recommendations for guideline 2 (intake of processed foods), the WCRF/AICR scoring method from Turati et al., 2020 was used. This method has pre-defined cutoff values for the recommended levels of intake for each nutrient that is expected to have clinical significance. Based on their self-reported levels of intake of added sugar, saturated fat, and sodium in processed foods, participants receive adherence scores for each dietary domain of 0 (failure to meet recommendation), 0.5 (partially meeting recommended levels), or 1 (fully meeting recommended level of intake). Guideline 2 total adherence score (range 0-1) was calculated as the average of adherence scores for consumption of added sugar, saturated fat, and sodium in processed foods.

To calculate adherence to the NCI dietary recommendations for guideline 3 (intake of red and processed meats), the WCRF/AICR scoring method from Turati et al., 2020 was used. This method has pre-defined cutoff values for the recommended levels of intake for each nutrient that is expected to have clinical significance. Based on their self-reported levels of intake of red and processed meats, participants receive adherence scores for each dietary domain of 0 (failure to meet recommendation), 0.5 (partially meeting recommended levels), or 1 (fully meeting recommended level of intake). Guideline 3 total adherence score (range 0-1) was calculated as the average of adherence scores for red and processed meat intake.

To calculate adherence to the NCI dietary recommendations for guideline 4 (intake of sugar sweetened beverages), the WCRF/AICR scoring method from Turati et al., 2020 was used. This method has pre-defined cutoff values for the recommended levels of intake for each nutrient that is expected to have clinical significance. Based on their self-reported levels of intake of sugar sweetened beverages, participants receive an adherence score of 0 (failure to meet recommendation), 0.5 (partially meeting recommended levels), or 1 (fully meeting recommended level of intake). Guideline 4 total adherence score (range 0-1) was calculated as adherence score for intake of sugar sweetened beverages.

To measure grocery purchasing, an API transferred participants' item-level food purchases to the study database whenever they shopped at designated grocery stores (Wegmans, ShopRite, Target, and Walmart). Each item in the shopping trip was linked to an existing nutrition databases (e.g., FNDDS) to categorize it into a NCI relevant category and to pull nutritional information. Fruit purchases were calculated by tabulating the percent of fruit purchases divided by total purchases. Scores range from 0-100%.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

To measure grocery purchasing, an API transferred participants' item-level food purchases to the study database whenever they shopped at designated grocery stores (Wegmans, ShopRite, Target, and Walmart). Each item in the shopping trip was linked to an existing nutrition databases (e.g., FNDDS) to categorize it into a NCI relevant category and to pull nutritional information. Vegetable purchases were calculated by tabulating the percent of fruit purchases divided by total purchases. Scores range from 0-100%.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

To measure grocery purchasing, an API transferred participants' item-level food purchases to the study database whenever they shopped at designated grocery stores (Wegmans, ShopRite, Target, and Walmart). Each item in the shopping trip was linked to an existing nutrition databases (e.g., FNDDS) to categorize it into a NCI relevant category and to pull nutritional information. Meat purchases were calculated by tabulating the percent of meat purchases divided by total purchases. Scores range from 0-100%.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

To measure grocery purchasing, an API transferred participants' item-level food purchases to the study database whenever they shopped at designated grocery stores (Wegmans, ShopRite, Target, and Walmart). Each item in the shopping trip was linked to an existing nutrition databases (e.g., FNDDS) to categorize it into a NCI relevant category and to pull nutritional information. Grains/beans purchases were calculated by tabulating the percent of grains and beans purchases divided by total purchases. Scores range from 0-100%.This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

To measure grocery purchasing, an API transferred participants' item-level food purchases to the study database whenever they shopped at designated stores (Wegmans, ShopRite, Target, and Walmart). Each item in the shopping trip was linked to an existing nutrition databases (e.g., FNDDS) to categorize it into a NCI relevant category and to pull nutritional information. Sugar-sweetened beverage (SSB) purchases were calculated by tabulating the percent of SSB purchases divided by total purchases. Scores range from 0-100%.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

To measure grocery purchasing, an API transferred participants' item-level food purchases to the study database whenever they shopped at designated grocery stores (Wegmans, ShopRite, Target, and Walmart). Each item in the shopping trip was linked to an existing nutrition databases (e.g., FNDDS) to categorize it into a NCI relevant category and to pull nutritional information. Sweets purchases were calculated by tabulating the percent of meat purchases divided by total sweets purchases. Scores range from 0-100%.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

To measure grocery purchasing, an API transferred participants' item-level food purchases to the study database whenever they shopped at designated grocery stores (Wegmans, ShopRite, Target, and Walmart). Each item in the shopping trip was linked to an existing nutrition databases (e.g., FNDDS) to categorize it into a NCI relevant category and to pull nutritional information. Snack purchases were calculated by tabulating the percent of snack purchases divided by total purchases. Scores range from 0-100%.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

To measure grocery purchasing, an API transferred participants' item-level food purchases to the study database whenever they shopped at designated grocery stores (Wegmans, ShopRite, Target, and Walmart). Each item in the shopping trip was linked to an existing nutrition databases (e.g., FNDDS) to categorize it into a NCI relevant category and to pull nutritional information. Processed food purchases were calculated by tabulating the percent of processed food purchases divided by total purchases. Scores range from 0-100%.This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

Location triggered messaging - ON, Benefits of change - ON, Coach monitoring - OFF, Household support - OFF.

Location triggered messaging - ON, Benefits of change - ON, Coach monitoring - ON, Household support - OFF.

Location triggered messaging - ON, Benefits of change - ON, Coach monitoring - ON, Household support - ON.

Location triggered messaging - ON, Benefits of change - ON, Coach monitoring - OFF, Household support - ON.

Location triggered messaging - ON, Benefits of change - OFF, Coach monitoring - ON, Household support - ON.

Location triggered messaging - ON, Benefits of change - OFF, Coach monitoring - OFF, Household support - ON.

Location triggered messaging - ON, Benefits of change - OFF, Coach monitoring - ON, Household support - OFF.

Location triggered messaging - OFF, Benefits of change - ON, Coach monitoring - OFF, Household support - OFF.

Location triggered messaging - OFF, Benefits of change - ON, Coach monitoring - ON, Household support - OFF.

Location triggered messaging - OFF, Benefits of change - ON, Coach monitoring - ON, Household support - ON.

Location triggered messaging - OFF, Benefits of change - ON, Coach monitoring - OFF, Household support - ON.

Location triggered messaging - OFF, Benefits of change - OFF, Coach monitoring - ON Household support - OFF.

Location triggered messaging - OFF, Benefits of change - OFF, Coach monitoring - OFF, Household support - ON.

Location triggered messaging - OFF, Benefits of change - OFF, Coach monitoring - ON, Household support - ON.

Total of all reporting groups

In Wave 1, the primary outcome will be change from 0-20 weeks in dietary intake of the major food categories from the American Institute for Cancer Research guidelines. Self-reported dietary recalls will be completed for three days at baseline and week 20 using the Automated Self Administered 24-hour food recall (ASA-24). This table reports the change in number of cups of fruits and vegetables per day.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

Participants were randomly assigned to whether each of the four conditions were "on" (location triggered messaging, coach monitoring, benefits of change, and household report). Change across time for each outcome are reported for each condition and for the full sample.

The Ns reported here are smaller than the full trial sample because only Wave 1 participants completed the ASA-24. Results here only include Wave 1 participants.

In Wave 1, the primary outcome will be change from 0-20 weeks in dietary intake of the major food categories from the American Institute for Cancer Research guidelines. Self-reported dietary recalls will be completed for three days at baseline and week 20 using the Automated Self Administered 24-hour food recall (ASA-24). This table reports the change in ounces of red meat consumed per week.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

Participants were randomly assigned to whether each of the four conditions were "on" (location triggered messaging, coach monitoring, benefits of change, and household report). Change across time for each outcome are reported for each condition and for the full sample.

The Ns reported here are smaller than the full trial sample because only Wave 1 participants completed the ASA-24. Results here only include Wave 1 participants.

In Wave 1, the primary outcome will be change from 0-20 weeks in dietary intake of the major food categories from the American Institute for Cancer Research guidelines. Self-reported dietary recalls will be completed for three days at baseline and week 20 using the Automated Self Administered 24-hour food recall (ASA-24). This table reports the change in ounces of processed meat consumed per week.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

Participants were randomly assigned to whether each of the four conditions were "on" (location triggered messaging, coach monitoring, benefits of change, and household report). Change across time for each outcome are reported for each condition and for the full sample.

The Ns reported here are smaller than the full trial sample because only Wave 1 participants completed the ASA-24. Results here only include Wave 1 participants.

In Wave 1, the primary outcome will be change from 0-20 weeks in dietary intake of the major food categories from the American Institute for Cancer Research guidelines. Self-reported dietary recalls will be completed for three days at baseline and week 20 using the Automated Self Administered 24-hour food recall (ASA-24). This table reports the change in grams of added sugars consumed per day.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

Participants were randomly assigned to whether each of the four conditions were "on" (location triggered messaging, coach monitoring, benefits of change, and household report). Change across time for each outcome are reported for each condition and for the full sample.

The Ns reported here are smaller than the full trial sample because only Wave 1 participants completed the ASA-24. Results here only include Wave 1 participants.

In Wave 1, the primary outcome will be change from 0-20 weeks in dietary intake of the major food categories from the American Institute for Cancer Research guidelines. Self-reported dietary recalls will be completed for three days at baseline and week 20 using the Automated Self Administered 24-hour food recall (ASA-24). This table reports the change in mg of sodium consumed per day.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

Participants were randomly assigned to whether each of the four conditions were "on" (location triggered messaging, coach monitoring, benefits of change, and household report). Change across time for each outcome are reported for each condition and for the full sample.

The Ns reported here are smaller than the full trial sample because only Wave 1 participants completed the ASA-24. Results here only include Wave 1 participants.

In Wave 1, the primary outcome will be change from 0-20 weeks in dietary intake of the major food categories from the American Institute for Cancer Research guidelines. Self-reported dietary recalls will be completed for three days at baseline and week 20 using the Automated Self Administered 24-hour food recall (ASA-24). This table reports the change in grams of saturated fat consumed per day.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

Participants were randomly assigned to whether each of the four conditions were "on" (location triggered messaging, coach monitoring, benefits of change, and household report). Change across time for each outcome are reported for each condition and for the full sample.

The Ns reported here are smaller than the full trial sample because only Wave 1 participants completed the ASA-24. Results here only include Wave 1 participants.

In Wave 1, the primary outcome will be change from 0-20 weeks in dietary intake of the major food categories from the American Institute for Cancer Research guidelines. Self-reported dietary recalls will be completed for three days at baseline and week 20 using the Automated Self Administered 24-hour food recall (ASA-24). This table reports the change in oz of sugar sweetened beverages consumed per day.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

Participants were randomly assigned to whether each of the four conditions were "on" (location triggered messaging, coach monitoring, benefits of change, and household report). Change across time for each outcome are reported for each condition and for the full sample.

The Ns reported here are smaller than the full trial sample because only Wave 1 participants completed the ASA-24. Results here only include Wave 1 participants.

In Wave 2, the Diet History Questionnaire (DHQ-III) will be used instead of food recalls to assess dietary intake. This table reports the change in grams of fiber per day.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

Participants were randomly assigned to whether each of the four conditions were "on" (location triggered messaging, coach monitoring, benefits of change, and household report). Change across time for each outcome are reported for each condition and for the full sample.

The Ns reported here are smaller than the full trial sample because only Wave 2 participants completed the DHQ-III. Results here only include Wave 2 participants.

In Wave 2, the Diet History Questionnaire (DHQ-III) will be used instead of food recalls to assess dietary intake. This table reports the change in cups of fruits and vegetables per day.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

Participants were randomly assigned to whether each of the four conditions were "on" (location triggered messaging, coach monitoring, benefits of change, and household report). Change across time for each outcome are reported for each condition and for the full sample.

The Ns reported here are smaller than the full trial sample because only Wave 2 participants completed the DHQ-III. Results here only include Wave 2 participants.

In Wave 2, the Diet History Questionnaire (DHQ-III) will be used instead of food recalls to assess dietary intake. This table reports the change in oz of red meat consumed per week.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

Participants were randomly assigned to whether each of the four conditions were "on" (location triggered messaging, coach monitoring, benefits of change, and household report). Change across time for each outcome are reported for each condition and for the full sample.

The Ns reported here are smaller than the full trial sample because only Wave 2 participants completed the DHQ-III. Results here only include Wave 2 participants.

In Wave 2, the Diet History Questionnaire (DHQ-III) will be used instead of food recalls to assess dietary intake. This table reports the change in oz of processed meat consumed per week.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

Participants were randomly assigned to whether each of the four conditions were "on" (location triggered messaging, coach monitoring, benefits of change, and household report). Change across time for each outcome are reported for each condition and for the full sample.

The Ns reported here are smaller than the full trial sample because only Wave 2 participants completed the DHQ-III. Results here only include Wave 2 participants.

In Wave 2, the Diet History Questionnaire (DHQ-III) will be used instead of food recalls to assess dietary intake. This table reports the change in grams of added sugars consumed per day.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

Participants were randomly assigned to whether each of the four conditions were "on" (location triggered messaging, coach monitoring, benefits of change, and household report). Change across time for each outcome are reported for each condition and for the full sample.

The Ns reported here are smaller than the full trial sample because only Wave 2 participants completed the DHQ-III. Results here only include Wave 2 participants.

In Wave 2, the Diet History Questionnaire (DHQ-III) will be used instead of food recalls to assess dietary intake. This table reports the change in mg of sodium consumed per day.

This pilot study used a factorial design, with 16 conditions and typically 4 participants per condition. The study was designed with the intent to collapse across conditions for all analyses, such that the 8 conditions in which a given factor is ON would be compared to the 8 conditions in which that factor was OFF. Participants are counted multiple times in the table, such that the total number of participants in any ON/OFF condition pair will equal the full sample.

Participants were randomly assigned to whether each of the four conditions were "on" (location triggered messaging, coach monitoring, benefits of change, and household report). Change across time for each outcome are reported for each condition and for the full sample.

The Ns reported here are smaller than the full trial sample because only Wave 2 participants completed the DHQ-III. Results here only include Wave 2 participants.