Common use of The MEPS Sampling Process and Response Rates Clause in Contracts

The MEPS Sampling Process and Response Rates. An Overview Generally, a sample representing about three-eighths of the NHIS responding households is made available for use in MEPS. This was the case for Panels 17 and 18. A household consists of one or more MEPS reporting units (RUs). A member of an RU is asked to provide MEPS data for all RU members. A multi-person RU consists of people who are related (based on the MEPS definition of a family—see the discussion of family weights below for more details). A single person RU is a person living in a household with no relatives present. A subsample of the responding NHIS households is drawn for MEPS interviewing. Because the MEPS subsampling has to be done soon after NHIS responding households are identified, a small percentage of the NHIS households initially characterized as NHIS respondents are later classified as nonrespondents for the purposes of NHIS data analysis. This actually serves to increase the overall MEPS response rate slightly since the percentage of NHIS households eligible for MEPS is slightly larger than the NHIS household-level response rate and some NHIS nonresponding households do participate in MEPS. However, as a result, these NHIS nonrespondents who are MEPS participants have no NHIS data available to link to MEPS data. Once the MEPS sample is selected from among the NHIS households characterized as NHIS respondents, RUs representing students living in student housing or consisting entirely of military personnel are dropped from the sample. For the NHIS, college students living in student housing are sampled independently of their families. For MEPS, such students are identified through the sample selection of their parents’ RU. Removing from MEPS those college students found in college housing sampled for the NHIS eliminates the opportunity of multiple chances of selection for MEPS for these students. Military personnel not living in the same RU as civilians are ineligible for MEPS. After such exclusions, all RUs associated with households selected from among those identified as NHIS responding households are then fielded in the first round of MEPS. Table 3-1 shows the three informational components just discussed in Rows A, B, and C. Row A indicates the percentage of NHIS households eligible for MEPS. Row B indicates the number of NHIS households sampled for MEPS. Row C indicates the number of sampled households actually fielded for MEPS (after dropping the students and military members discussed above). From these numbers unweighted response rates have been computed, allowing an assessment of the propensity to participate among those RUs actually sampled for MEPS. It should be noted that the sample distribution of RUs has a disproportionately high percentage of several minorities and MEPS response rates vary by race/ethnicity. Table 3-1. Unweighted response rates for point-in-time file (Panel 18 Round 1/Panel 17 Round 3) Panel 17 Panel 18 2013 Combined A. Percentage of NHIS sample eligible for MEPS* 82.9% 78.0% B. Number of households sampled from the NHIS 9,700 9,700 C. Number of Households sampled from the NHIS and fielded for MEPS 9,676 9,685 D. Round 1 – Number of RUs eligible for interviewing 10,386 10,357 E. Round 1 – Number of RUs with completed interviews 8,121 7,683 F. Round 2 – Number of RUs eligible for interviewing 8,359 G. Round 2 – Number of RUs with completed interviews 7,874 H. Round 3 – Number of RUs eligible for interviewing 8,049 I. Round 3 – Number of RUs with completed interviews 7,663 Overall response rates through the Spring of 2013 P17: A x (E/D) x (G/F) x (I/H) P18: A x (E/D) Combined: .487 x P17 +.513 x P18 *among the panels and quarters of the NHIS allocated to MEPS, the percentage of households that were considered to be NHIS respondents at the time the MEPS sample was selected 58.1% (Panel 17) through Round 3) 57.9% (Panel 18) through Round 1) 58.0% When an RU is visited for a round of data collection, changes in RU membership are identified. Such changes include RU members who have moved to another location in the U.S., thus creating a new RU to be interviewed for MEPS, as well as student RUs. Thus, the number of RUs known to be eligible for MEPS interviewing in a given round can only be determined after data collection is fully completed. The ratio of the number of RUs completing the MEPS interview in a given round to the number of RUs characterized as eligible to complete the interview for that round represents the “conditional” response rate for that round expressed as a proportion. It is “conditional” in that it pertains to the set of RUs characterized as eligible for MEPS specifically for that round, and thus is “conditioned” on prior participation rather than representing the overall response rate through that round. For example, in Table 3.1, for Panel 18, Round 1 the ratio of 7,683 (Row E) to 10,357 (Row D) multiplied by 100 is 74.2 and represents the unweighted response rate for the round conditioned on the set of RUs characterized as eligible for MEPS for Round 1, expressed as a percentage. Multiplying the percentage of the NHIS sample eligible for MEPS (Row A) by the product of the ratios for a consecutive set of MEPS rounds beginning with round one produces the overall response rate through the last MEPS round specified. (It should be noted that the number of RUs with completed interviews is slightly higher than the number of RUs receiving family weights. RUs receiving family weights must satisfy additional criteria such as participation throughout the entire period of eligibility by all key, in-scope RU members and the requirement that the RU reference person must be key.) The overall response rate for the combined sample of Panel 18, Round 1 and Panel 17, Round 3 is obtained by taking the sum of the products of the relative sample sizes and the corresponding overall panel response rates. The details of the process of computing the overall response rate is described below in Section 3.1.4.

The MEPS Sampling Process and Response Rates. An Overview Generally, a sample representing about three-eighths of the NHIS responding households is made available for use in MEPS. This was the case for Panels 17 13 and 18. A household consists of one or more MEPS reporting units (RUs). A member of an RU is asked to provide MEPS data for all RU members. A multi-person RU consists of people who are related (based on the MEPS definition of a family—see the discussion of family weights below for more details). A single person RU is a person living in a household with no relatives present14. A subsample of the responding NHIS households is drawn for MEPS interviewing. Because the MEPS subsampling has to be done soon after NHIS responding households are identified, a small percentage of the NHIS households initially characterized as NHIS respondents are later classified as nonrespondents for the purposes of NHIS data analysis. This actually serves to increase the overall MEPS response rate slightly since the percentage of NHIS households eligible for MEPS is slightly larger than the NHIS household-level response rate and some NHIS nonresponding households do participate in MEPS. However, as a result, these NHIS nonrespondents who are MEPS participants have no NHIS data available to link to MEPS data. Once the MEPS sample is selected from among the NHIS households characterized as NHIS respondents, RUs representing students living in student housing or consisting entirely of military personnel are dropped from the sample. For the NHIS, college students living in student housing are sampled independently of their families. For MEPS, such students are identified through the sample selection of their parents’ RU. Removing from MEPS those college students found in college housing sampled for the NHIS eliminates the opportunity of multiple chances of selection for MEPS for these students. Military personnel not living in the same RU as civilians are ineligible for MEPS. After such exclusions, all RUs associated with households selected from among those identified as NHIS responding households are then fielded in the first round of MEPS. Table 3-1 shows the three informational components just discussed in Rows A, B, and C. Row A indicates the percentage of NHIS households eligible for MEPS. Row B indicates the number of NHIS households sampled for MEPS. Row C indicates the number of sampled households actually fielded for MEPS (after dropping the students and military members discussed above). From these numbers unweighted response rates have been computed, allowing an assessment of the propensity to participate among those RUs actually sampled for MEPS. It should be noted that the sample distribution of RUs has Those actually sampled for MEPS include a disproportionately high percentage number of members of several minorities and MEPS response rates vary by race/ethnicityas well as those “predicted to be poor” (based on a statistical model) compared to the distribution of such groups in the general population. The “predicted to be poor” were oversampled for Panel 13 only. Table 3-1. Unweighted response rates for pointPoint-in-time Time file (Panel 18 14 Round 1/Panel 17 13, Round 3) Panel 17 13 Panel 18 2013 14 2009 Combined A. Percentage of NHIS sample eligible for MEPS* 82.9MEPS 87.4% 78.085.2% B. Number of households sampled from the NHIS 9,700 9,703 9,700 C. Number of Households sampled from the NHIS and fielded for MEPS 9,676 9,685 9,688 9,672 D. Round 1 – Number of RUs eligible for interviewing 10,386 10,357 10,325 10,227 E. Round 1 – Number of RUs with completed interviews 8,121 7,683 8,017 7,650 F. Round 2 – Number of RUs eligible for interviewing 8,359 8,252 G. Round 2 – Number of RUs with completed interviews 7,874 7,809 H. Round 3 – Number of RUs eligible for interviewing 8,049 7,982 I. Round 3 – Number of RUs with completed interviews 7,663 7,684 Overall response rates through the Spring of 2013 P172009 61.8% 63.7% 62.8% P13: A x (E/D) x (G/F) x (I/H) P18P14: A x (E/D) Combined: .487 .495 x P17 P13 +.513 .505 x P18 *among the panels and quarters of the NHIS allocated to MEPS, the percentage of households that were considered to be NHIS respondents at the time the MEPS sample was selected 58.1% P14 (Panel 17) 13 through Round 3) 57.9% (Panel 1814) through Round 1) 58.0% When an RU is visited for a round of data collection, changes in RU membership are identified. Such changes include RU members who have moved to another location in the U.S., thus creating a new RU to be interviewed for MEPS, as well as student RUs. Thus, the number of RUs known to be eligible for MEPS interviewing in a given round can only be determined after data collection is fully completed. The ratio of the number of RUs completing the MEPS interview in a given round to the number of RUs characterized as eligible to complete the interview for that round represents the “conditional” response rate for that round expressed as a proportion. It is “conditional” in that it pertains to the set of RUs characterized as eligible for MEPS specifically for that round, and thus is “conditioned” on prior participation rather than representing the overall response rate through that round. For example, in Table 3.1, for Panel 1814, Round 1 the ratio of 7,683 7,650 (Row E) to 10,357 10,227 (Row D) multiplied by 100 is 74.2 74.8 and represents the unweighted response rate for the round conditioned on the set of RUs characterized as eligible for MEPS for Round 1, expressed as a percentage. Multiplying the percentage of the NHIS sample eligible for MEPS (Row row A) by the product of the ratios for a consecutive set of MEPS rounds beginning with round one produces the overall response rate through the last MEPS round specified. (It should be noted that the number of RUs with completed interviews is slightly higher than the number of RUs receiving family weights. RUs receiving family weights must satisfy additional criteria such as participation throughout the entire period of eligibility by all key, in-scope inscope RU members and the requirement that the RU reference person must be key.) The overall response rate for the combined sample of Panel 1814, Round 1 and Panel 1713, Round 3 is obtained by taking the sum of the products of the relative sample sizes and the corresponding overall panel response rates. The details Panel 13, Round 3 represents about 49.5 percent of the process of computing combined sample size, while Panel 14, Round 1 represents approximately 50.5 percent. Thus, the overall combined response rate is described below in Section 3.1.4has been computed as .495 multiplied by the Panel 13 response rate through Round 3 plus .505 multiplied by the Panel 14 response rate through Round 1. As a point of information, for Panel 13 an experiment has been conducted with respect to the effectiveness of various levels of incentives to increase response rates and/or the quality of the data collected. Analyses of the results of the experiment are currently being conducted.

The MEPS Sampling Process and Response Rates. An Overview GenerallyFor most MEPS panels, a sample representing about three-eighths of the NHIS responding households is made available for use in MEPS. This was the case for Panels 17 both MEPS Panel 20 and 18. A household consists of one or more MEPS reporting units (RUs). A member of an RU is asked to provide MEPS data for all RU members. A multi-person RU consists of people who are related (based on the MEPS definition of a family—see the discussion of family weights below for more details). A single person RU is a person living in a household with no relatives present. A subsample of the responding NHIS households is drawn for MEPS interviewingPanel 21. Because the MEPS subsampling has to be done soon after NHIS responding households are identified, a small percentage of the NHIS households initially characterized as NHIS respondents are later classified as nonrespondents for the purposes of NHIS data analysis. This actually serves to increase the overall MEPS response rate slightly since the percentage of NHIS households eligible designated for use in MEPS (all those characterized initially as respondents from the NHIS panels and quarters used by MEPS for a given year) is slightly larger than the final NHIS household-level response rate and some NHIS nonresponding households do participate in MEPS. However, as a result, these NHIS nonrespondents who are MEPS participants have no NHIS data available to link to with MEPS data. Once the MEPS sample is selected from among the NHIS households characterized as NHIS respondents, RUs representing students living in student housing or consisting entirely of military personnel are dropped deleted from the sample. For the NHIS, college students living in student housing are sampled independently of from their families. For MEPS, such students are identified through the sample selection of their parents’ RU. Removing from MEPS those college students found in college housing sampled for the NHIS eliminates the opportunity of multiple chances of selection for MEPS for these students. Military personnel not living in the same RU as civilians are ineligible for MEPS. After such exclusions, all RUs associated with households selected from among those identified as NHIS responding households are then fielded in the first round of MEPS. Table 3-1 3.1 shows the three informational components just discussed in Rows A, B, and C. C the three informational components just discussed. Row A indicates the percentage of NHIS households eligible for MEPS. Row B indicates the number of NHIS households sampled for MEPS. Row C indicates the number of sampled households actually fielded for MEPS (after dropping the students and military members discussed above). From these numbers unweighted Note that all response rates have been computed, allowing an assessment of the propensity to participate among those RUs actually sampled for MEPS. It should be noted that the sample distribution of RUs has a disproportionately high percentage of several minorities and MEPS response rates vary by race/ethnicity. Table 3-1. Unweighted response rates for point-in-time file (Panel 18 Round 1/Panel 17 Round 3) Panel 17 Panel 18 2013 Combined A. Percentage of NHIS sample eligible for MEPS* 82.9% 78.0% B. Number of households sampled from the NHIS 9,700 9,700 C. Number of Households sampled from the NHIS and fielded for MEPS 9,676 9,685 D. Round 1 – Number of RUs eligible for interviewing 10,386 10,357 E. Round 1 – Number of RUs with completed interviews 8,121 7,683 F. Round 2 – Number of RUs eligible for interviewing 8,359 G. Round 2 – Number of RUs with completed interviews 7,874 H. Round 3 – Number of RUs eligible for interviewing 8,049 I. Round 3 – Number of RUs with completed interviews 7,663 Overall response rates through the Spring of 2013 P17: A x (E/D) x (G/F) x (I/H) P18: A x (E/D) Combined: .487 x P17 +.513 x P18 *among the panels and quarters of the NHIS allocated to MEPS, the percentage of households that were considered to be NHIS respondents at the time the MEPS sample was selected 58.1% (Panel 17) through Round 3) 57.9% (Panel 18) through Round 1) 58.0% When an RU is visited for a round of data collection, changes in RU membership discussed here are identified. Such changes include RU members who have moved to another location in the U.S., thus creating a new RU to be interviewed for MEPS, as well as student RUs. Thus, the number of RUs known to be eligible for MEPS interviewing in a given round can only be determined after data collection is fully completed. The ratio of the number of RUs completing the MEPS interview in a given round to the number of RUs characterized as eligible to complete the interview for that round represents the “conditional” response rate for that round expressed as a proportion. It is “conditional” in that it pertains to the set of RUs characterized as eligible for MEPS specifically for that round, and thus is “conditioned” on prior participation rather than representing the overall response rate through that round. For example, in Table 3.1, for Panel 18, Round 1 the ratio of 7,683 (Row E) to 10,357 (Row D) multiplied by 100 is 74.2 and represents the unweighted response rate for the round conditioned on the set of RUs characterized as eligible for MEPS for Round 1, expressed as a percentage. Multiplying the percentage of the NHIS sample eligible for MEPS (Row A) by the product of the ratios for a consecutive set of MEPS rounds beginning with round one produces the overall response rate through the last MEPS round specified. (It should be noted that the number of RUs with completed interviews is slightly higher than the number of RUs receiving family weights. RUs receiving family weights must satisfy additional criteria such as participation throughout the entire period of eligibility by all key, in-scope RU members and the requirement that the RU reference person must be keyunweighted.) The overall response rate for the combined sample of Panel 18, Round 1 and Panel 17, Round 3 is obtained by taking the sum of the products of the relative sample sizes and the corresponding overall panel response rates. The details of the process of computing the overall response rate is described below in Section 3.1.4.

The MEPS Sampling Process and Response Rates. An Overview Generally, a sample representing about three-eighths of the NHIS responding households is made available for use in MEPS. This was the case for Panels 17 and 18. A household consists of one or more MEPS reporting units (RUs). A member of an RU is asked to provide MEPS data for all RU members. A multi-person RU consists of people who are related (based on the MEPS definition of a family—see the discussion of family weights below for more details). A single person RU is a person living in a household with no relatives present. A subsample of the responding NHIS these households is then drawn for MEPS interviewing. Because the MEPS subsampling has to be done soon after NHIS responding households are identified, a small percentage of the NHIS households initially characterized as NHIS respondents are later classified as nonrespondents for the purposes of NHIS data analysis. This actually serves to increase the overall MEPS response rate slightly since the percentage of NHIS households eligible for MEPS is slightly larger than the NHIS household-level response rate and some NHIS nonresponding households do participate in MEPS. However, as a result, these NHIS nonrespondents who are MEPS participants have no NHIS data available to link to incorporate into analyses with MEPS data. Once the MEPS sample is selected from among the NHIS households characterized as NHIS respondents, RUs representing students living in student housing or consisting entirely of military personnel are dropped from the sample. For the NHIS, college students living in student housing are sampled independently of from their families. For MEPS, such students are identified through the sample selection of their parents’ ' RU. Removing from MEPS those college students found in college housing sampled for the NHIS eliminates the opportunity of multiple chances of selection for MEPS for these students. Military personnel not living in the same RU as civilians are ineligible for MEPS. After such exclusions, all RUs associated with households selected from among those identified as NHIS responding households are then fielded in the first round of MEPS. Table 3-1 3.1 shows the these three informational components just discussed in Rows A, B, and C. Row A indicates the percentage of NHIS households eligible for MEPS. Row B indicates the number of NHIS households sampled for MEPS. Row C indicates the number of sampled households actually fielded for MEPS (after dropping the students and military members discussed above). From these numbers unweighted response rates have been computed, allowing an assessment of the propensity to participate among those RUs actually sampled for MEPS. It should be noted that the sample distribution of RUs has a disproportionately high percentage of several minorities and MEPS response rates vary by race/ethnicity. Table 3-1. Unweighted response rates for point-in-time file (Panel 18 Round 1/Panel 17 Round 3) Panel 17 Panel 18 2013 Combined A. Percentage of NHIS sample eligible for MEPS* 82.9% 78.0% B. Number of households sampled from the NHIS 9,700 9,700 C. Number of Households sampled from the NHIS and fielded for MEPS 9,676 9,685 D. Round 1 – Number of RUs eligible for interviewing 10,386 10,357 E. Round 1 – Number of RUs with completed interviews 8,121 7,683 F. Round 2 – Number of RUs eligible for interviewing 8,359 G. Round 2 – Number of RUs with completed interviews 7,874 H. Round 3 – Number of RUs eligible for interviewing 8,049 I. Round 3 – Number of RUs with completed interviews 7,663 Overall response rates through the Spring of 2013 P17: A x (E/D) x (G/F) x (I/H) P18: A x (E/D) Combined: .487 x P17 +.513 x P18 *among the panels and quarters of the NHIS allocated to MEPS, the percentage of households that were considered to be NHIS respondents at the time the MEPS sample was selected 58.1% (Panel 17) through Round 3) 57.9% (Panel 18) through Round 1) 58.0% When an RU is visited for a round of data collection, changes in RU membership are identified. Such changes include RU members who have moved to another location in the U.S., thus creating a new RU to be interviewed for MEPS, as well as student RUs. Thus, the number of RUs known to be eligible for MEPS interviewing in a given round can only be determined after data collection is fully completed. The ratio of the number of RUs completing the MEPS interview in a given round to the number of RUs characterized as eligible to complete the interview for that round represents the “conditional” response rate for that round expressed as a proportion. It is “conditional” in that it pertains to the set of RUs characterized as eligible for MEPS specifically for that round, and thus is “conditioned” on prior participation rather than representing the overall response rate through that round. For example, in Table 3.1, for Panel 18, Round 1 the ratio of 7,683 (Row E) to 10,357 (Row D) multiplied by 100 is 74.2 and represents the unweighted response rate for the round conditioned on the set of RUs characterized as eligible for MEPS for Round 1, expressed as a percentage. Multiplying the percentage of the NHIS sample eligible for MEPS (Row A) by the product of the ratios for a consecutive set of MEPS rounds beginning with round one produces the overall response rate through the last MEPS round specified. (It should be noted that the number of RUs with completed interviews is slightly higher than the number of RUs receiving family weights. RUs receiving family weights must satisfy additional criteria such as participation throughout the entire period of eligibility by all key, in-scope RU members and the requirement that the RU reference person must be key.) The overall response rate for the combined sample of Panel 18, Round 1 and Panel 17, Round 3 is obtained by taking the sum of the products of the relative sample sizes and the corresponding overall panel response rates. The details of the process of computing the overall response rate is described below in Section 3.1.4.

The MEPS Sampling Process and Response Rates. An Overview Generally, a sample representing about three-eighths of the NHIS responding households is made available for use in MEPS. This was the case for Panels 17 19 and 1820. A household consists of one or more MEPS reporting units (RUs). A member of an RU is asked to provide MEPS data for all RU members. A multi-person RU consists of people who are related (based on the MEPS definition of a family—see the discussion of family weights below for more details). A single person RU is a person living in a household with no relatives present. A subsample of the responding NHIS households is drawn for MEPS interviewing. Because the MEPS subsampling has to be done soon after NHIS responding households are identified, a small percentage of the NHIS households initially characterized as NHIS respondents are later classified as nonrespondents for the purposes of NHIS data analysis. This actually serves to increase the overall MEPS response rate slightly since the percentage of NHIS households eligible for MEPS is slightly larger than the NHIS household-level response rate and some NHIS nonresponding households do participate in MEPS. However, as a result, these NHIS nonrespondents who are MEPS participants have no NHIS data available to link to MEPS data. Once the MEPS sample is selected from among the NHIS households characterized as NHIS respondents, RUs representing students living in student housing or consisting entirely of military personnel are dropped from the sample. For the NHIS, college students living in student housing are sampled independently of their families. For MEPS, such students are identified through the sample selection of their parents’ RU. Removing from MEPS those college students found in college housing sampled for the NHIS eliminates the opportunity of multiple chances of selection for MEPS for these students. Military personnel not living in the same RU as civilians are ineligible for MEPS. After such exclusions, all RUs associated with households selected from among those identified as NHIS responding households are then fielded in the first round of MEPS. Table 3-1 shows the three informational components just discussed in Rows A, B, and C. Row A indicates the percentage of NHIS households eligible for MEPS. Row B indicates the number of NHIS households sampled for MEPS. Row C indicates the number of sampled households actually fielded for MEPS (after dropping the students and military members discussed above). From these numbers unweighted response rates have been computed, allowing an assessment of the propensity to participate among those RUs actually sampled for MEPS. It should be noted that the sample distribution of RUs has a disproportionately high percentage of several minorities and MEPS response rates vary by race/ethnicity. Table 3-1. Unweighted response rates for point-in-time file (Panel 18 20 Round 1/Panel 17 19 Round 3) Panel 17 19 Panel 18 2013 20 2015 Combined A. Percentage of NHIS sample eligible for MEPS* 82.976.2% 78.075.1% B. Number of households sampled from the NHIS 9,700 9,700 10,610 Panel 19 Panel 20 2015 Combined C. Number of Households sampled from the NHIS and fielded for MEPS 9,676 9,685 9,667 10,571 D. Round 1 – Number of RUs eligible for interviewing 10,386 10,357 10,355 11,318 E. Round 1 – Number of RUs with completed interviews 8,121 7,683 7,436 8,313 F. Round 2 – Number of RUs eligible for interviewing 8,359 7,678 G. Round 2 – Number of RUs with completed interviews 7,874 7,179 H. Round 3 – Number of RUs eligible for interviewing 8,049 7,343 I. Round 3 – Number of RUs with completed interviews 7,663 6,956 Overall response rates through the Spring of 2013 P172015 P19: A x (E/D) x (G/F) x (I/H) P18P20: A x (E/D) Combined: .487 .44 x P17 P19 +.513 .56 x P18 P20 *among the panels and quarters of the NHIS allocated to MEPS, the percentage of households that were considered to be NHIS respondents at the time the MEPS sample was selected 58.148.5% (Panel 1719) through Round 3) 57.955.1% (Panel 1820) through Round 1) 58.052.2% When an RU is visited for a round of data collection, changes in RU membership are identified. Such changes include RU members who have moved to another location in the U.S., thus creating a new RU to be interviewed for MEPS, as well as student RUs. Thus, the number of RUs known to be eligible for MEPS interviewing in a given round can only be determined after data collection is fully completed. The ratio of the number of RUs completing the MEPS interview in a given round to the number of RUs characterized as eligible to complete the interview for that round represents the “conditional” response rate for that round expressed as a proportion. It is “conditional” in that it pertains to the set of RUs characterized as eligible for MEPS specifically for that round, and thus is “conditioned” on prior participation rather than representing the overall response rate through that round. For example, in Table 3.13-1, for Panel 1820, Round 1 the ratio of 7,683 8,313 (Row E) to 10,357 11,318 (Row D) multiplied by 100 is 74.2 73.4 and represents the unweighted response rate for the round conditioned on the set of RUs characterized as eligible for MEPS for Round 1, expressed as a percentage. Multiplying the percentage of the NHIS sample eligible for MEPS (Row A) by the product of the ratios for a consecutive set of MEPS rounds beginning with round one produces the overall response rate through the last MEPS round specified. (It should be noted that the number of RUs with completed interviews is slightly higher than the number of RUs receiving family weights. RUs receiving family weights must satisfy additional criteria such as participation throughout the entire period of eligibility by all key, in-scope RU members and the requirement that the RU reference person must be key.) The overall response rate for the combined sample of Panel 1820, Round 1 and Panel 1719, Round 3 is obtained by taking the sum of the products of the relative sample sizes and the corresponding overall panel response rates. The details of the process of computing the overall response rate is described below in Section 3.1.4.

The MEPS Sampling Process and Response Rates. An Overview GenerallyOverview‌ For most MEPS panels, a sample representing about three-eighths of the NHIS responding households is made available for use in MEPS. This was the case for Panels 17 MEPS Panel 23, Panel 24, and 18Panel 25. A household consists of one or more For Panel 26, the sample made available for use in MEPS reporting units (RUs). A member of an RU is asked to provide MEPS data for all RU members. A multi-person RU consists of people who are related (based on the MEPS definition of a family—see the discussion of family weights below for more details). A single person RU is a person living in a household with no relatives present. A subsample represented about five- eighths of the NHIS responding NHIS households is drawn for MEPS interviewingthe reasons discussed in Section 3.1.4. Because the MEPS subsampling has to be done soon after NHIS responding households are identified, a small percentage of the NHIS households initially characterized as NHIS respondents are later classified as nonrespondents for the purposes of NHIS data analysis. This actually serves to increase the overall MEPS response rate slightly since the percentage of NHIS households eligible designated for use in MEPS (all those characterized initially as respondents from the NHIS panels and quarters used by MEPS for a given year) is slightly larger than the final NHIS household-level response rate and some NHIS nonresponding households do participate in MEPS. However, as a result, these NHIS nonrespondents who are MEPS participants have no NHIS data available to link to with MEPS data. Once the MEPS sample is selected from among the NHIS households characterized as NHIS respondents, RUs representing students living in student housing or consisting entirely of military personnel are dropped deleted from the sample. For the NHIS, college students living in student housing are sampled independently of their families. For MEPS, such students are identified through the sample selection of their parents’ RU. Removing from MEPS those college students found in college housing sampled for the NHIS eliminates the opportunity of multiple chances of selection for MEPS for these students. Military personnel not living in the same RU as civilians are ineligible for MEPS. After such these exclusions, all RUs associated with households selected from among those identified as NHIS responding households are then fielded in the first round of MEPS. Table 3-1 3.1 shows the three informational components just discussed in Rows A, B, and C. C the three informational components just discussed. Row A indicates the percentage of NHIS households eligible for MEPS. Row B indicates the number of NHIS households sampled for MEPS. Row C indicates the number of sampled households actually fielded for MEPS (after dropping the students and military members discussed aboveabove and a small number of NHIS households sampled in error). From these numbers unweighted Note that all response rates have been computed, allowing an assessment of the propensity to participate among those RUs actually sampled for MEPS. It should be noted that the sample distribution of RUs has a disproportionately high percentage of several minorities and MEPS response rates vary by race/ethnicity. Table 3-1. Unweighted response rates for point-in-time file (Panel 18 Round 1/Panel 17 Round 3) Panel 17 Panel 18 2013 Combined A. Percentage of NHIS sample eligible for MEPS* 82.9% 78.0% B. Number of households sampled from the NHIS 9,700 9,700 C. Number of Households sampled from the NHIS and fielded for MEPS 9,676 9,685 D. Round 1 – Number of RUs eligible for interviewing 10,386 10,357 E. Round 1 – Number of RUs with completed interviews 8,121 7,683 F. Round 2 – Number of RUs eligible for interviewing 8,359 G. Round 2 – Number of RUs with completed interviews 7,874 H. Round 3 – Number of RUs eligible for interviewing 8,049 I. Round 3 – Number of RUs with completed interviews 7,663 Overall response rates through the Spring of 2013 P17: A x (E/D) x (G/F) x (I/H) P18: A x (E/D) Combined: .487 x P17 +.513 x P18 *among the panels and quarters of the NHIS allocated to MEPS, the percentage of households that were considered to be NHIS respondents at the time the MEPS sample was selected 58.1% (Panel 17) through Round 3) 57.9% (Panel 18) through Round 1) 58.0% When an RU is visited for a round of data collection, changes in RU membership discussed here are identified. Such changes include RU members who have moved to another location in the U.S., thus creating a new RU to be interviewed for MEPS, as well as student RUs. Thus, the number of RUs known to be eligible for MEPS interviewing in a given round can only be determined after data collection is fully completed. The ratio of the number of RUs completing the MEPS interview in a given round to the number of RUs characterized as eligible to complete the interview for that round represents the “conditional” response rate for that round expressed as a proportion. It is “conditional” in that it pertains to the set of RUs characterized as eligible for MEPS specifically for that round, and thus is “conditioned” on prior participation rather than representing the overall response rate through that round. For example, in Table 3.1, for Panel 18, Round 1 the ratio of 7,683 (Row E) to 10,357 (Row D) multiplied by 100 is 74.2 and represents the unweighted response rate for the round conditioned on the set of RUs characterized as eligible for MEPS for Round 1, expressed as a percentage. Multiplying the percentage of the NHIS sample eligible for MEPS (Row A) by the product of the ratios for a consecutive set of MEPS rounds beginning with round one produces the overall response rate through the last MEPS round specified. (It should be noted that the number of RUs with completed interviews is slightly higher than the number of RUs receiving family weights. RUs receiving family weights must satisfy additional criteria such as participation throughout the entire period of eligibility by all key, in-scope RU members and the requirement that the RU reference person must be keyunweighted.) The overall response rate for the combined sample of Panel 18, Round 1 and Panel 17, Round 3 is obtained by taking the sum of the products of the relative sample sizes and the corresponding overall panel response rates. The details of the process of computing the overall response rate is described below in Section 3.1.4.

The MEPS Sampling Process and Response Rates. An Overview Generally, a sample representing about three-eighths of the NHIS responding households is made available for use in MEPS. This was the case for Panels 17 and 18. A household consists of one or more MEPS reporting units (RUs). A member of an RU is asked to provide MEPS data for all RU members. A multi-person RU consists of people who are related (based on the MEPS definition of a family—see the discussion of family weights below for more details). A single person RU is a person living in a household with no relatives present. A subsample of the responding NHIS these households is then drawn for MEPS interviewing. Because the MEPS subsampling has to be done soon after NHIS responding households are identified, a small percentage of the NHIS households initially characterized as NHIS respondents are later classified as nonrespondents for the purposes of NHIS data analysis. This actually serves to increase the overall MEPS response rate slightly since the percentage of NHIS households eligible for MEPS is slightly larger than the NHIS household-level response rate and some NHIS nonresponding households do participate in MEPS. However, as a result, these NHIS nonrespondents who are MEPS participants have no NHIS data available to link to MEPS data. Once the MEPS sample is selected from among the NHIS households characterized as NHIS respondents, RUs representing students living in student housing or consisting entirely of military personnel are dropped from the sample. For the NHIS, college students living in student housing are sampled independently of their families. For MEPS, such students are identified through the sample selection of their parents’ ' RU. Removing from MEPS those college students found in college housing sampled for the NHIS eliminates the opportunity of multiple chances of selection for MEPS for these students. Military personnel not living in the same RU as civilians are ineligible for MEPS. After such exclusions, all RUs associated with households selected from among those identified as NHIS responding households are then fielded in the first round of MEPS. Table 3-1 3.1 shows the three informational components just discussed in Rows A, B, and C. Row A indicates the percentage of NHIS households eligible for MEPS. Row B indicates the number of NHIS households sampled for MEPS. Row C indicates the number of sampled households actually fielded for MEPS (after dropping the students and military members discussed above). From these numbers unweighted response rates have been computed, allowing an assessment of the propensity to participate among those RUs actually sampled for MEPS. It should be noted that the sample distribution of RUs has a disproportionately high percentage of several minorities and MEPS response rates vary by race/ethnicity. Table 3-1. Unweighted response 3.1 Response rates for pointPoint-in-time Time file (Panel 18 11 Round 1/Panel 17 10, Round 3) Panel 17 10 Panel 18 2013 11 2006 Combined A. Percentage of NHIS sample eligible for MEPS* 82.9MEPS 87.9% 78.087.3% B. Number of households sampled from the NHIS 9,700 9,700 8,546 9,464 C. Number of Households sampled from the NHIS and fielded for MEPS 9,676 9,685 8,505 9,434 D. Round 1 – Number of RUs eligible for interviewing 10,386 10,357 9,076 9,972 Panel 10 Panel 11 2006 Combined E. Round 1 – Number of RUs with completed interviews 8,121 7,683 7,170 7,585 F. Round 2 – Number of RUs eligible for interviewing 8,359 7,381 G. Round 2 – Number of RUs with completed interviews 7,874 6,940 H. Round 3 – Number of RUs eligible for interviewing 8,049 7,056 I. Round 3 – Number of RUs with completed interviews 7,663 6,727 Overall response rates through the Spring of 2013 P172006 62.2% (Panel 10 through round 3) 66.4% (Panel 11) through round 1) 64.5% P10: A x (E/D) x (G/F) x (I/H) P18P11: A x (E/D) Combined: .487 .46 x P17 P10 +.513 .54 x P18 *among the panels and quarters of the NHIS allocated to MEPS, the percentage of households that were considered to be NHIS respondents at the time the MEPS sample was selected 58.1% (Panel 17) through Round 3) 57.9% (Panel 18) through Round 1) 58.0% P11 When an RU is visited for a round of data collection, changes in RU membership are identified. Such changes include RU members who have moved to another location in the U.S., thus creating a new RU to be interviewed for MEPS, as well as student RUs. Thus, the number of RUs known to be eligible for MEPS interviewing in a given round can only be determined after data collection is fully completed. The ratio of the number of RUs completing the MEPS interview in a given round to the number of RUs characterized as eligible to complete the interview for that round represents the “"conditional” " response rate for that round round, expressed as a proportion. It is “"conditional” " in that it pertains to the set of RUs characterized as eligible for MEPS specifically for that round, and thus is “"conditioned” " on prior participation rather than representing the overall response rate through that round. For example, in Table 3.1, for Panel 1811, Round 1 the ratio of 7,683 7,585 (Row E) to 10,357 9,972 (Row D) multiplied by 100 is 74.2 76.1 and represents the unweighted response rate for the round conditioned on the set of RUs characterized as eligible for MEPS for Round 1, expressed as a percentage. Multiplying the percentage of the NHIS sample eligible for MEPS (Row row A) by the product of the ratios for a consecutive set of MEPS rounds beginning with round one produces the overall response rate through the last MEPS round specified. (It should be noted that the number of RUs with completed interviews is slightly higher than the number of RUs receiving family weights. RUs receiving family weights must satisfy additional criteria such as participation throughout the entire period of eligibility by all key, in-scope inscope RU members and the requirement that the RU reference person must be key.) The overall response rate for the combined sample of Panel 1811, Round 1 and Panel 1710, Round 3 is obtained by taking the sum of the products of the relative sample sizes and the corresponding overall panel response rates. The details Panel 10, Round 3 represents about 46 percent of the process of computing combined sample size, while Panel 11, Round 1 represents approximately 54 percent. Thus, the overall combined response rate is described below in Section 3.1.4has been computed as .46 multiplied by the Panel 10 response rate through Round 3 plus .54 multiplied by the Panel 11 response rate through Round 1.

The MEPS Sampling Process and Response Rates. An Overview Generally, a sample representing about three-eighths of the NHIS responding households is made available for use in MEPS. This was the case for Panels 17 15 and 18. A household consists of one or more MEPS reporting units (RUs). A member of an RU is asked to provide MEPS data for all RU members. A multi-person RU consists of people who are related (based on the MEPS definition of a family—see the discussion of family weights below for more details). A single person RU is a person living in a household with no relatives present16. A subsample of the responding NHIS households is drawn for MEPS interviewing. Because the MEPS subsampling has to be done soon after NHIS responding households are identified, a small percentage of the NHIS households initially characterized as NHIS respondents are later classified as nonrespondents for the purposes of NHIS data analysis. This actually serves to increase the overall MEPS response rate slightly since the percentage of NHIS households eligible for MEPS is slightly larger than the NHIS household-level response rate and some NHIS nonresponding households do participate in MEPS. However, as a result, these NHIS nonrespondents who are MEPS participants have no NHIS data available to link to MEPS data. Once the MEPS sample is selected from among the NHIS households characterized as NHIS respondents, RUs representing students living in student housing or consisting entirely of military personnel are dropped from the sample. For the NHIS, college students living in student housing are sampled independently of their families. For MEPS, such students are identified through the sample selection of their parents’ RU. Removing from MEPS those college students found in college housing sampled for the NHIS eliminates the opportunity of multiple chances of selection for MEPS for these students. Military personnel not living in the same RU as civilians are ineligible for MEPS. After such exclusions, all RUs associated with households selected from among those identified as NHIS responding households are then fielded in the first round of MEPS. Table 3-1 shows the three informational components just discussed in Rows A, B, and C. Row A indicates the percentage of NHIS households eligible for MEPS. Row B indicates the number of NHIS households sampled for MEPS. Row C indicates the number of sampled households actually fielded for MEPS (after dropping the students and military members discussed above). From these numbers unweighted response rates have been computed, allowing an assessment of the propensity to participate among those RUs actually sampled for MEPS. It should be noted that the sample distribution of RUs has Those actually sampled for MEPS include a disproportionately high percentage number of members of several minorities and MEPS minorities. As alluded to above, the subsampling of some Panel 16 households characterized as “interim nonrespondents” in 11 PSUs has resulted in the need to determine response rates vary by race/ethnicitythat reflect the fact that the contribution of the sampled “interim nonrespondents” is doubled while those not sampled do not count at all. The resulting response rates, found in Table 3-1 are still characterized as “unweighted” since they only reflect one component of the sampling for MEPS. Table 3-1. Unweighted response rates for pointPoint-in-time Time file (Panel 18 16 Round 1/Panel 17 15, Round 3) Panel 17 15 Panel 18 2013 16 2011 Combined A. Percentage of NHIS sample eligible for MEPS* 82.984.0% 78.080.6% B. Number of households sampled from the NHIS 9,700 9,700 C. Number of Households sampled from the NHIS and fielded for MEPS 9,676 9,685 D. Round 1 – Number of RUs eligible for interviewing 10,386 10,357 E. Round 1 – Number of RUs with completed interviews 8,121 7,683 F. Round 2 – Number of RUs eligible for interviewing 8,359 G. Round 2 – Number of RUs with completed interviews 7,874 H. Round 3 – Number of RUs eligible for interviewing 8,049 I. Round 3 – Number of RUs with completed interviews 7,663 Overall response rates through the Spring of 2013 P17: A x (E/D) x (G/F) x (I/H) P18: A x (E/D) Combined: .487 x P17 +.513 x P18 *among the panels and quarters of the NHIS allocated to MEPS, the percentage of households that were considered to be NHIS respondents at the time the MEPS sample was selected 58.1% (Panel 17) through Round 3) 57.9% (Panel 18) through Round 1) 58.0% When an RU is visited for a round of data collection, changes in RU membership are identified. Such changes include RU members who have moved to another location in the U.S., thus creating a new RU to be interviewed for MEPS, as well as student RUs. Thus, the number of RUs known to be eligible for MEPS interviewing in a given round can only be determined after data collection is fully completed. The ratio of the number of RUs completing the MEPS interview in a given round to the number of RUs characterized as eligible to complete the interview for that round represents the “conditional” response rate for that round expressed as a proportion. It is “conditional” in that it pertains to the set of RUs characterized as eligible for MEPS specifically for that round, and thus is “conditioned” on prior participation rather than representing the overall response rate through that round. For example, in Table 3.1, for Panel 18, Round 1 the ratio of 7,683 (Row E) to 10,357 (Row D) multiplied by 100 is 74.2 and represents the unweighted response rate for the round conditioned on the set of RUs characterized as eligible for MEPS for Round 1, expressed as a percentage. Multiplying the percentage of the NHIS sample eligible for MEPS (Row A) by the product of the ratios for a consecutive set of MEPS rounds beginning with round one produces the overall response rate through the last MEPS round specified. (It should be noted that the number of RUs with completed interviews is slightly higher than the number of RUs receiving family weights. RUs receiving family weights must satisfy additional criteria such as participation throughout the entire period of eligibility by all key, in-scope RU members and the requirement that the RU reference person must be key.) The overall response rate for the combined sample of Panel 18, Round 1 and Panel 17, Round 3 is obtained by taking the sum of the products of the relative sample sizes and the corresponding overall panel response rates. The details of the process of computing the overall response rate is described below in Section 3.1.4.8,750 10,180

The MEPS Sampling Process and Response Rates. An Overview Generally, a sample representing about three-eighths of the NHIS responding households is made available for use in MEPS. This was the case for Panels 17 and 18Panel 13. A household consists For Panel 12 about one quarter of one or more MEPS reporting units (RUs). A member of an RU is asked to provide MEPS data the NHIS responding households were sampled for all RU members. A multi-person RU consists of people who are related (based on the MEPS definition of a family—see the discussion of family weights below for more details). A single person RU is a person living in a household with no relatives presentMEPS. A subsample of the responding NHIS households is drawn for MEPS interviewing. Because the MEPS subsampling has to be done soon after NHIS responding households are identified, a small percentage of the NHIS households initially characterized as NHIS respondents are later classified as nonrespondents for the purposes of NHIS data analysis. This actually serves to increase the overall MEPS response rate slightly since the percentage of NHIS households eligible for MEPS is slightly larger than the NHIS household-level response rate and some NHIS nonresponding households do participate in MEPS. However, as a result, these NHIS nonrespondents who are MEPS participants have no NHIS data available to link to MEPS data. Once the MEPS sample is selected from among the NHIS households characterized as NHIS respondents, RUs representing students living in student housing or consisting entirely of military personnel are dropped from the sample. For the NHIS, college students living in student housing are sampled independently of their families. For MEPS, such students are identified through the sample selection of their parents’ RU. Removing from MEPS those college students found in college housing sampled for the NHIS eliminates the opportunity of multiple chances of selection for MEPS for these students. Military personnel not living in the same RU as civilians are ineligible for MEPS. After such exclusions, all RUs associated with households selected from among those identified as NHIS responding households are then fielded in the first round of MEPS. Table 3-1 shows the three informational components just discussed in Rows A, B, and C. Row A indicates the percentage of NHIS households eligible for MEPS. Row B indicates the number of NHIS households sampled for MEPS. Row C indicates the number of sampled households actually fielded for MEPS (after dropping the students and military members discussed above). From these numbers unweighted response rates have been computed, allowing an assessment of the propensity to participate among those RUs actually sampled for MEPS. It should be noted that the sample distribution of RUs has Those actually sampled for MEPS include a disproportionately high percentage number of members of several minorities and MEPS response rates vary by race/ethnicityas well as those “predicted to be poor” (based on a statistical model) compared to the distribution of such groups in the general population. Table 3-1. Unweighted response rates for pointPoint-in-time Time file (Panel 18 13 Round 1/Panel 17 12, Round 3) Panel 17 12 Panel 18 2013 13 2008 Combined A. Percentage of NHIS sample eligible for MEPS* 82.9MEPS 88.1% 78.0% B. Number of households sampled from the NHIS 9,700 9,700 C. Number of Households sampled from the NHIS and fielded for MEPS 9,676 9,685 D. Round 1 – Number of RUs eligible for interviewing 10,386 10,357 E. Round 1 – Number of RUs with completed interviews 8,121 7,683 F. Round 2 – Number of RUs eligible for interviewing 8,359 G. Round 2 – Number of RUs with completed interviews 7,874 H. Round 3 – Number of RUs eligible for interviewing 8,049 I. Round 3 – Number of RUs with completed interviews 7,663 Overall response rates through the Spring of 2013 P17: A x (E/D) x (G/F) x (I/H) P18: A x (E/D) Combined: .487 x P17 +.513 x P18 *among the panels and quarters of the NHIS allocated to MEPS, the percentage of households that were considered to be NHIS respondents at the time the MEPS sample was selected 58.1% (Panel 17) through Round 3) 57.9% (Panel 18) through Round 1) 58.0% When an RU is visited for a round of data collection, changes in RU membership are identified. Such changes include RU members who have moved to another location in the U.S., thus creating a new RU to be interviewed for MEPS, as well as student RUs. Thus, the number of RUs known to be eligible for MEPS interviewing in a given round can only be determined after data collection is fully completed. The ratio of the number of RUs completing the MEPS interview in a given round to the number of RUs characterized as eligible to complete the interview for that round represents the “conditional” response rate for that round expressed as a proportion. It is “conditional” in that it pertains to the set of RUs characterized as eligible for MEPS specifically for that round, and thus is “conditioned” on prior participation rather than representing the overall response rate through that round. For example, in Table 3.1, for Panel 18, Round 1 the ratio of 7,683 (Row E) to 10,357 (Row D) multiplied by 100 is 74.2 and represents the unweighted response rate for the round conditioned on the set of RUs characterized as eligible for MEPS for Round 1, expressed as a percentage. Multiplying the percentage of the NHIS sample eligible for MEPS (Row A) by the product of the ratios for a consecutive set of MEPS rounds beginning with round one produces the overall response rate through the last MEPS round specified. (It should be noted that the number of RUs with completed interviews is slightly higher than the number of RUs receiving family weights. RUs receiving family weights must satisfy additional criteria such as participation throughout the entire period of eligibility by all key, in-scope RU members and the requirement that the RU reference person must be key.) The overall response rate for the combined sample of Panel 18, Round 1 and Panel 17, Round 3 is obtained by taking the sum of the products of the relative sample sizes and the corresponding overall panel response rates. The details of the process of computing the overall response rate is described below in Section 3.1.4.87.4%

The MEPS Sampling Process and Response Rates. An Overview Generally, a sample representing about three-eighths of the NHIS responding households is made available for use in MEPS. This was the case for Panels 17 and 18. A household consists of one or more MEPS reporting units (RUs). A member of an RU is asked to provide MEPS data for all RU members. A multi-person RU consists of people who are related (based on the MEPS definition of a family—see the discussion of family weights below for more details). A single person RU is a person living in a household with no relatives present. A subsample of the responding NHIS these households is then drawn for MEPS interviewing. Because the MEPS subsampling has to be done soon after NHIS responding households are identified, a small percentage of the NHIS households initially characterized as NHIS respondents are later classified as nonrespondents for the purposes of NHIS data analysis. This actually serves to increase the overall MEPS response rate slightly since the percentage of NHIS households eligible for MEPS is slightly larger than the NHIS household-level response rate and some NHIS nonresponding households do participate in MEPS. However, as a result, these NHIS nonrespondents who are MEPS participants have no NHIS data available to link to incorporate into analyses with MEPS data. Once the MEPS sample is selected from among the NHIS households characterized as NHIS respondents, RUs representing students living in student housing or consisting entirely of military personnel are dropped from the sample. For the NHIS, college students living in student housing are sampled independently of their families. For MEPS, such students are identified through the sample selection of their parents’ ' RU. Removing from MEPS those college students found in college housing sampled for the NHIS eliminates the opportunity of multiple chances of selection for MEPS for these students. Military personnel not living in the same RU as civilians are ineligible for MEPS. After such exclusions, all RUs associated with households selected from among those identified as NHIS responding households are then fielded in the first round of MEPS. Table 3-1 3.1 shows the three informational components just discussed in Rows A, B, and C. Row A indicates the percentage of NHIS households eligible for MEPS. Row B indicates the number of NHIS households sampled for MEPS. Row C indicates the number of sampled households actually fielded for MEPS (after dropping the students and military members discussed above). From these numbers unweighted response rates have been computed, allowing an assessment of the propensity to participate among those RUs actually sampled for MEPS. It should be noted that the sample distribution of RUs has a disproportionately high percentage of several minorities and MEPS response rates vary by race/ethnicity. Table 3-1. Unweighted response rates for point-in-time file (Panel 18 Round 1/Panel 17 Round 3) Panel 17 Panel 18 2013 Combined A. Percentage of NHIS sample eligible for MEPS* 82.9% 78.0% B. Number of households sampled from the NHIS 9,700 9,700 C. Number of Households sampled from the NHIS and fielded for MEPS 9,676 9,685 D. Round 1 – Number of RUs eligible for interviewing 10,386 10,357 E. Round 1 – Number of RUs with completed interviews 8,121 7,683 F. Round 2 – Number of RUs eligible for interviewing 8,359 G. Round 2 – Number of RUs with completed interviews 7,874 H. Round 3 – Number of RUs eligible for interviewing 8,049 I. Round 3 – Number of RUs with completed interviews 7,663 Overall response rates through the Spring of 2013 P17: A x (E/D) x (G/F) x (I/H) P18: A x (E/D) Combined: .487 x P17 +.513 x P18 *among the panels and quarters of the NHIS allocated to MEPS, the percentage of households that were considered to be NHIS respondents at the time the MEPS sample was selected 58.1% (Panel 17) through Round 3) 57.9% (Panel 18) through Round 1) 58.0% When an RU is visited for a round of data collection, changes in RU membership are identified. Such changes include RU members who have moved to another location in the U.S., thus creating a new RU to be interviewed for MEPS, as well as student RUs. Thus, the number of RUs known to be eligible for MEPS interviewing in a given round can only be determined after data collection is fully completed. The ratio of the number of RUs completing the MEPS interview in a given round to the number of RUs characterized as eligible to complete the interview for that round represents the “conditional” response rate for that round expressed as a proportion. It is “conditional” in that it pertains to the set of RUs characterized as eligible for MEPS specifically for that round, and thus is “conditioned” on prior participation rather than representing the overall response rate through that round. For example, in Table 3.1, for Panel 18, Round 1 the ratio of 7,683 (Row E) to 10,357 (Row D) multiplied by 100 is 74.2 and represents the unweighted response rate for the round conditioned on the set of RUs characterized as eligible for MEPS for Round 1, expressed as a percentage. Multiplying the percentage of the NHIS sample eligible for MEPS (Row A) by the product of the ratios for a consecutive set of MEPS rounds beginning with round one produces the overall response rate through the last MEPS round specified. (It should be noted that the number of RUs with completed interviews is slightly higher than the number of RUs receiving family weights. RUs receiving family weights must satisfy additional criteria such as participation throughout the entire period of eligibility by all key, in-scope RU members and the requirement that the RU reference person must be key.) The overall response rate for the combined sample of Panel 18, Round 1 and Panel 17, Round 3 is obtained by taking the sum of the products of the relative sample sizes and the corresponding overall panel response rates. The details of the process of computing the overall response rate is described below in Section 3.1.4.

The MEPS Sampling Process and Response Rates. An Overview Generally, a sample representing about three-eighths of the NHIS responding households is made available for use in MEPS. This was the case for Panels 17 14 and 18. A household consists of one or more MEPS reporting units (RUs). A member of an RU is asked to provide MEPS data for all RU members. A multi-person RU consists of people who are related (based on the MEPS definition of a family—see the discussion of family weights below for more details). A single person RU is a person living in a household with no relatives present15. A subsample of the responding NHIS households is drawn for MEPS interviewing. Because the MEPS subsampling has to be done soon after NHIS responding households are identified, a small percentage of the NHIS households initially characterized as NHIS respondents are later classified as nonrespondents for the purposes of NHIS data analysis. This actually serves to increase the overall MEPS response rate slightly since the percentage of NHIS households eligible for MEPS is slightly larger than the NHIS household-level response rate and some NHIS nonresponding households do participate in MEPS. However, as a result, these NHIS nonrespondents who are MEPS participants have no NHIS data available to link to MEPS data. Once the MEPS sample is selected from among the NHIS households characterized as NHIS respondents, RUs representing students living in student housing or consisting entirely of military personnel are dropped from the sample. For the NHIS, college students living in student housing are sampled independently of their families. For MEPS, such students are identified through the sample selection of their parents’ RU. Removing from MEPS those college students found in college housing sampled for the NHIS eliminates the opportunity of multiple chances of selection for MEPS for these students. Military personnel not living in the same RU as civilians are ineligible for MEPS. After such exclusions, all RUs associated with households selected from among those identified as NHIS responding households are then fielded in the first round of MEPS. Table 3-1 shows the three informational components just discussed in Rows A, B, and C. Row A indicates the percentage of NHIS households eligible for MEPS. Row B indicates the number of NHIS households sampled for MEPS. Row C indicates the number of sampled households actually fielded for MEPS (after dropping the students and military members discussed above). From these numbers unweighted response rates have been computed, allowing an assessment of the propensity to participate among those RUs actually sampled for MEPS. It should be noted that the sample distribution of RUs has Those actually sampled for MEPS include a disproportionately high percentage number of members of several minorities and MEPS response rates vary by race/ethnicityminorities. Table 3-1. Unweighted response rates for pointPoint-in-time Time file (Panel 18 15 Round 1/Panel 17 14, Round 3) Panel 17 14 Panel 18 2013 15 2010 Combined A. Percentage of NHIS sample eligible for MEPS* 82.9MEPS 85.2% 78.084.0% B. Number of households sampled from the NHIS 9,700 9,700 8,750 C. Number of Households sampled from the NHIS and fielded for MEPS 9,676 9,685 9,672 8,726 D. Round 1 – Number of RUs eligible for interviewing 10,386 10,357 10,227 9,258 E. Round 1 – Number of RUs with completed interviews 8,121 7,683 7,650 6,802 F. Round 2 – Number of RUs eligible for interviewing 8,359 7,883 G. Round 2 – Number of RUs with completed interviews 7,874 7,239 H. Round 3 – Number of RUs eligible for interviewing 8,049 7,359 I. Round 3 – Number of RUs with completed interviews 7,663 6,980 Overall response rates through the Spring of 2013 P172010 55.5% 61.7% 58.6% P14: A x (E/D) x (G/F) x (I/H) P18P15: A x (E/D) Combined: .487 .495 x P17 P14 +.513 .505 x P18 *among the panels and quarters of the NHIS allocated to MEPS, the percentage of households that were considered to be NHIS respondents at the time the MEPS sample was selected 58.1% P15 (Panel 17) 14 through Round 3) 57.9% (Panel 1815) through Round 1) 58.0% When an RU is visited for a round of data collection, changes in RU membership are identified. Such changes include RU members who have moved to another location in the U.S., thus creating a new RU to be interviewed for MEPS, as well as student RUs. Thus, the number of RUs known to be eligible for MEPS interviewing in a given round can only be determined after data collection is fully completed. The ratio of the number of RUs completing the MEPS interview in a given round to the number of RUs characterized as eligible to complete the interview for that round represents the “conditional” response rate for that round expressed as a proportion. It is “conditional” in that it pertains to the set of RUs characterized as eligible for MEPS specifically for that round, and thus is “conditioned” on prior participation rather than representing the overall response rate through that round. For example, in Table 3.1, for Panel 1815, Round 1 the ratio of 7,683 6,802 (Row E) to 10,357 9,258 (Row D) multiplied by 100 is 74.2 73.5 and represents the unweighted response rate for the round conditioned on the set of RUs characterized as eligible for MEPS for Round 1, expressed as a percentage. Multiplying the percentage of the NHIS sample eligible for MEPS (Row row A) by the product of the ratios for a consecutive set of MEPS rounds beginning with round one produces the overall response rate through the last MEPS round specified. (It should be noted that the number of RUs with completed interviews is slightly higher than the number of RUs receiving family weights. RUs receiving family weights must satisfy additional criteria such as participation throughout the entire period of eligibility by all key, in-scope inscope RU members and the requirement that the RU reference person must be key.) The overall response rate for the combined sample of Panel 1815, Round 1 and Panel 1714, Round 3 is obtained by taking the sum of the products of the relative sample sizes and the corresponding overall panel response rates. The details Panel 14, Round 3 represents about 49.5 percent of the process of computing combined sample size, while Panel 15, Round 1 represents approximately 50.5 percent. Thus, the overall combined response rate is described below in Section 3.1.4has been computed as .495 multiplied by the Panel 14 response rate through Round 3 plus .505 multiplied by the Panel 15 response rate through Round 1. As a point of information, for Panel 14 an experiment has been conducted with respect to the effectiveness of various levels of incentives to increase response rates and/or the quality of the data collected. Analyses of the results of the experiment are currently being conducted.

The MEPS Sampling Process and Response Rates. An Overview Generally, a sample representing about three-eighths of the NHIS responding households is made available for use in MEPS. This was the case for Panels 17 18 and 1819. A household consists of one or more MEPS reporting units (RUs). A member of an RU is asked to provide MEPS data for all RU members. A multi-person RU consists of people who are related (based on the MEPS definition of a family—see the discussion of family weights below for more details). A single person RU is a person living in a household with no relatives present. A subsample of the responding NHIS households is drawn for MEPS interviewing. Because the MEPS subsampling has to be done soon after NHIS responding households are identified, a small percentage of the NHIS households initially characterized as NHIS respondents are later classified as nonrespondents for the purposes of NHIS data analysis. This actually serves to increase the overall MEPS response rate slightly since the percentage of NHIS households eligible for MEPS is slightly larger than the NHIS household-level response rate and some NHIS nonresponding households do participate in MEPS. However, as a result, these NHIS nonrespondents who are MEPS participants have no NHIS data available to link to MEPS data. Once the MEPS sample is selected from among the NHIS households characterized as NHIS respondents, RUs representing students living in student housing or consisting entirely of military personnel are dropped from the sample. For the NHIS, college students living in student housing are sampled independently of their families. For MEPS, such students are identified through the sample selection of their parents’ RU. Removing from MEPS those college students found in college housing sampled for the NHIS eliminates the opportunity of multiple chances of selection for MEPS for these students. Military personnel not living in the same RU as civilians are ineligible for MEPS. After such exclusions, all RUs associated with households selected from among those identified as NHIS responding households are then fielded in the first round of MEPS. Table 3-1 shows the three informational components just discussed in Rows A, B, and C. Row A indicates the percentage of NHIS households eligible for MEPS. Row B indicates the number of NHIS households sampled for MEPS. Row C indicates the number of sampled households actually fielded for MEPS (after dropping the students and military members discussed above). From these numbers unweighted response rates have been computed, allowing an assessment of the propensity to participate among those RUs actually sampled for MEPS. It should be noted that the sample distribution of RUs has a disproportionately high percentage of several minorities and MEPS response rates vary by race/ethnicity. Table 3-1. Unweighted response rates for point-in-time file (Panel 18 19 Round 1/Panel 17 18 Round 3) Panel 17 18 Panel 18 2013 19 2014 Combined A. Percentage of NHIS sample eligible for MEPS* 82.978.0% 78.076.2% B. Number of households sampled from the NHIS 9,700 9,700 Panel 18 Panel 19 2014 Combined C. Number of Households sampled from the NHIS and fielded for MEPS 9,676 9,685 9,667 D. Round 1 – Number of RUs eligible for interviewing 10,386 10,357 10,355 E. Round 1 – Number of RUs with completed interviews 8,121 7,683 7,436 F. Round 2 – Number of RUs eligible for interviewing 8,359 7,971 G. Round 2 – Number of RUs with completed interviews 7,874 7,402 H. Round 3 – Number of RUs eligible for interviewing 8,049 7,622 I. Round 3 – Number of RUs with completed interviews 7,663 7,205 Overall response rates through the Spring of 2013 P172014 P18: A x (E/D) x (G/F) x (I/H) P18P19: A x (E/D) Combined: .487 x P17 +.513 .48 x P18 +.52 x P19 *among the panels and quarters of the NHIS allocated to MEPS, the percentage of households that were considered to be NHIS respondents at the time the MEPS sample was selected 58.1% (Panel 17) through Round 3) 57.950.8% (Panel 18) through Round 3) 54.7% (Panel 19) through Round 1) 58.052.8% When an RU is visited for a round of data collection, changes in RU membership are identified. Such changes include RU members who have moved to another location in the U.S., thus creating a new RU to be interviewed for MEPS, as well as student RUs. Thus, the number of RUs known to be eligible for MEPS interviewing in a given round can only be determined after data collection is fully completed. The ratio of the number of RUs completing the MEPS interview in a given round to the number of RUs characterized as eligible to complete the interview for that round represents the “conditional” response rate for that round expressed as a proportion. It is “conditional” in that it pertains to the set of RUs characterized as eligible for MEPS specifically for that round, and thus is “conditioned” on prior participation rather than representing the overall response rate through that round. For example, in Table 3.1, for Panel 1819, Round 1 the ratio of 7,683 7,436 (Row E) to 10,357 10,355 (Row D) multiplied by 100 is 74.2 71.8 and represents the unweighted response rate for the round conditioned on the set of RUs characterized as eligible for MEPS for Round 1, expressed as a percentage. Multiplying the percentage of the NHIS sample eligible for MEPS (Row A) by the product of the ratios for a consecutive set of MEPS rounds beginning with round one produces the overall response rate through the last MEPS round specified. (It should be noted that the number of RUs with completed interviews is slightly higher than the number of RUs receiving family weights. RUs receiving family weights must satisfy additional criteria such as participation throughout the entire period of eligibility by all key, in-scope RU members and the requirement that the RU reference person must be key.) The overall response rate for the combined sample of Panel 1819, Round 1 and Panel 1718, Round 3 is obtained by taking the sum of the products of the relative sample sizes and the corresponding overall panel response rates. The details of the process of computing the overall response rate is described below in Section 3.1.4.

The MEPS Sampling Process and Response Rates. An Overview Generally, a sample representing about three-eighths of the NHIS responding households is are made available for use in MEPS. This was the case for Panels 17 and 18. A household consists of one or more MEPS reporting units (RUs). A member of an RU is asked to provide MEPS data for all RU members. A multi-person RU consists of people who are related (based on the MEPS definition of a family—see the discussion of family weights below for more details). A single person RU is a person living in a household with no relatives present. A subsample of the responding NHIS these households is then drawn for MEPS interviewing. Because the MEPS subsampling has to be done very soon after NHIS responding households are identified, a small percentage of the NHIS households initially characterized as NHIS respondents are later classified as nonrespondents for the purposes of NHIS data analysis. This actually serves to increase the overall MEPS response rate slightly since the percentage of NHIS households eligible for MEPS is slightly larger than the NHIS household-level response rate and some NHIS nonresponding households do participate in MEPS. However, as a result, for these NHIS nonrespondents who there are MEPS participants have no NHIS data available to link to incorporate into analyses with MEPS data. Once the MEPS sample is selected from among the NHIS households characterized as NHIS respondents, RUs representing students living in student housing or consisting entirely of military personnel are dropped from the sample. For the NHIS, college students living in student housing are sampled independently of from their families. For MEPS, such students are identified through the sample selection of their parents’ ' RU. Removing from MEPS those college students found in college housing sampled for the NHIS eliminates the opportunity of multiple chances of selection for MEPS for these students. Military personnel not living in the same RU as civilians are ineligible for MEPS. After such exclusions, all RUs associated with households selected from among those identified as NHIS responding households are then fielded in the first round of MEPS. Table 3-1 3.1 shows the these three informational components just discussed in Rows A, B, and C. Row A indicates the percentage of NHIS households eligible for MEPS. Row B indicates the number of NHIS households sampled for MEPS. Row C indicates the number of sampled households actually fielded for MEPS (after dropping the students and military members discussed above). From these numbers unweighted response rates have been computed, allowing an assessment of the propensity to participate among those RUs actually sampled for MEPS. It should be noted that the sample distribution of RUs has a disproportionately high percentage of several minorities and MEPS response rates vary by race/ethnicity. Table 3-1. Unweighted response rates for point-in-time file (Panel 18 Round 1/Panel 17 Round 3) Panel 17 Panel 18 2013 Combined A. Percentage of NHIS sample eligible for MEPS* 82.9% 78.0% B. Number of households sampled from the NHIS 9,700 9,700 C. Number of Households sampled from the NHIS and fielded for MEPS 9,676 9,685 D. Round 1 – Number of RUs eligible for interviewing 10,386 10,357 E. Round 1 – Number of RUs with completed interviews 8,121 7,683 F. Round 2 – Number of RUs eligible for interviewing 8,359 G. Round 2 – Number of RUs with completed interviews 7,874 H. Round 3 – Number of RUs eligible for interviewing 8,049 I. Round 3 – Number of RUs with completed interviews 7,663 Overall response rates through the Spring of 2013 P17: A x (E/D) x (G/F) x (I/H) P18: A x (E/D) Combined: .487 x P17 +.513 x P18 *among the panels and quarters of the NHIS allocated to MEPS, the percentage of households that were considered to be NHIS respondents at the time the MEPS sample was selected 58.1% (Panel 17) through Round 3) 57.9% (Panel 18) through Round 1) 58.0% When an RU is visited for a round of data collection, changes in RU membership are identified. Such changes include RU members who have moved to another location in the U.S., thus creating a new RU to be interviewed for MEPS, as well as student RUs. Thus, the number of RUs known to be eligible for MEPS interviewing in a given round can only be determined after data collection is fully completed. The ratio of the number of RUs completing the MEPS interview in a given round to the number of RUs characterized as eligible to complete the interview for that round represents the “conditional” response rate for that round expressed as a proportion. It is “conditional” in that it pertains to the set of RUs characterized as eligible for MEPS specifically for that round, and thus is “conditioned” on prior participation rather than representing the overall response rate through that round. For example, in Table 3.1, for Panel 18, Round 1 the ratio of 7,683 (Row E) to 10,357 (Row D) multiplied by 100 is 74.2 and represents the unweighted response rate for the round conditioned on the set of RUs characterized as eligible for MEPS for Round 1, expressed as a percentage. Multiplying the percentage of the NHIS sample eligible for MEPS (Row A) by the product of the ratios for a consecutive set of MEPS rounds beginning with round one produces the overall response rate through the last MEPS round specified. (It should be noted that the number of RUs with completed interviews is slightly higher than the number of RUs receiving family weights. RUs receiving family weights must satisfy additional criteria such as participation throughout the entire period of eligibility by all key, in-scope RU members and the requirement that the RU reference person must be key.) The overall response rate for the combined sample of Panel 18, Round 1 and Panel 17, Round 3 is obtained by taking the sum of the products of the relative sample sizes and the corresponding overall panel response rates. The details of the process of computing the overall response rate is described below in Section 3.1.4.

The MEPS Sampling Process and Response Rates. An Overview Generally, a sample representing about three-eighths of the NHIS responding households is made available for use in MEPS. This was the case for Panels 17 16 and 1817. A household consists of one or more MEPS reporting units (RUs). A member of an RU is asked to provide MEPS data for all RU members. A multi-person RU consists of people who are related (based on the MEPS definition of a family—see the discussion of family weights below for more details). A single person RU is a person living in a household with no relatives present. A subsample of the responding NHIS households is drawn for MEPS interviewing. Because the MEPS subsampling has to be done soon after NHIS responding households are identified, a small percentage of the NHIS households initially characterized as NHIS respondents are later classified as nonrespondents for the purposes of NHIS data analysis. This actually serves to increase the overall MEPS response rate slightly since the percentage of NHIS households eligible for MEPS is slightly larger than the NHIS household-level response rate and some NHIS nonresponding households do participate in MEPS. However, as a result, these NHIS nonrespondents who are MEPS participants have no NHIS data available to link to MEPS data. Once the MEPS sample is selected from among the NHIS households characterized as NHIS respondents, RUs representing students living in student housing or consisting entirely of military personnel are dropped from the sample. For the NHIS, college students living in student housing are sampled independently of their families. For MEPS, such students are identified through the sample selection of their parents’ RU. Removing from MEPS those college students found in college housing sampled for the NHIS eliminates the opportunity of multiple chances of selection for MEPS for these students. Military personnel not living in the same RU as civilians are ineligible for MEPS. After such exclusions, all RUs associated with households selected from among those identified as NHIS responding households are then fielded in the first round of MEPS. Table 3-1 shows the three informational components just discussed in Rows A, B, and C. Row A indicates the percentage of NHIS households eligible for MEPS. Row B indicates the number of NHIS households sampled for MEPS. Row C indicates the number of sampled households actually fielded for MEPS (after dropping the students and military members discussed above). From these numbers unweighted response rates have been computed, allowing an assessment of the propensity to participate among those RUs actually sampled for MEPS. It should be noted that the sample distribution of RUs has a disproportionately high percentage of several minorities and MEPS response rates vary by race/ethnicity. As alluded to above, the subsampling of some Panel 16 households characterized as “interim nonrespondents” in 11 PSUs has resulted in the need to determine Panel 16 response rates that reflect the fact that the contribution of the sampled “interim nonrespondents” is doubled while those not sampled do not count at all. The resulting response rates, found in Table 3-1 are still characterized as “unweighted” since they only reflect this one component of the sampling for MEPS. Table 3-1. Unweighted response rates for pointPoint-in-time Time file (Panel 18 17 Round 1/Panel 17 16, Round 3) Panel 16 Panel 17 Panel 18 2013 2012 Combined A. Percentage of NHIS sample eligible for MEPS* 80.6% 82.9% 78.0% B. Number of households sampled from the NHIS 9,700 9,700 C. Number of Households sampled from the NHIS and fielded for MEPS 9,676 9,685 D. Round 1 – Number of RUs eligible for interviewing 10,386 10,357 E. Round 1 – Number of RUs with completed interviews 8,121 7,683 F. Round 2 – Number of RUs eligible for interviewing 8,359 G. Round 2 – Number of RUs with completed interviews 7,874 H. Round 3 – Number of RUs eligible for interviewing 8,049 I. Round 3 – Number of RUs with completed interviews 7,663 Overall response rates through the Spring of 2013 P17: A x (E/D) x (G/F) x (I/H) P18: A x (E/D) Combined: .487 x P17 +.513 x P18 *among the panels and quarters of the NHIS allocated to MEPS, the percentage of households that were considered to be NHIS respondents at the time the MEPS sample was selected 58.1% (Panel 17) through Round 3) 57.9% (Panel 18) through Round 1) 58.0% When an RU is visited for a round of data collection, changes in RU membership are identified. Such changes include RU members who have moved to another location in the U.S., thus creating a new RU to be interviewed for MEPS, as well as student RUs. Thus, the number of RUs known to be eligible for MEPS interviewing in a given round can only be determined after data collection is fully completed. The ratio of the number of RUs completing the MEPS interview in a given round to the number of RUs characterized as eligible to complete the interview for that round represents the “conditional” response rate for that round expressed as a proportion. It is “conditional” in that it pertains to the set of RUs characterized as eligible for MEPS specifically for that round, and thus is “conditioned” on prior participation rather than representing the overall response rate through that round. For example, in Table 3.1, for Panel 18, Round 1 the ratio of 7,683 (Row E) to 10,357 (Row D) multiplied by 100 is 74.2 and represents the unweighted response rate for the round conditioned on the set of RUs characterized as eligible for MEPS for Round 1, expressed as a percentage. Multiplying the percentage of the NHIS sample eligible for MEPS (Row A) by the product of the ratios for a consecutive set of MEPS rounds beginning with round one produces the overall response rate through the last MEPS round specified. (It should be noted that the number of RUs with completed interviews is slightly higher than the number of RUs receiving family weights. RUs receiving family weights must satisfy additional criteria such as participation throughout the entire period of eligibility by all key, in-scope RU members and the requirement that the RU reference person must be key.) The overall response rate for the combined sample of Panel 18, Round 1 and Panel 17, Round 3 is obtained by taking the sum of the products of the relative sample sizes and the corresponding overall panel response rates. The details of the process of computing the overall response rate is described below in Section 3.1.4.%