Common use of Assignment of Weights Clause in Contracts

Assignment of Weights. If all key, in-scope members of a family responded to MEPS for their entire period of eligibility for Panel 6, Round 3 or for Panel 7, Round 1 and if the family had a key reference person, then that family received a family-level weight (WGTRU13>0). Reporting units consisting of an individual respondent who was both key and in-scope also received a family-level weight. These individual person units can be included or excluded from family-level analyses at the analyst’s discretion. Family-level weights were poststratified to figures obtained from the March 2002 CPS. The family-level poststratification reflected population distributions across family type (reference person married, spouse present; male reference person, no spouse present; female reference person, no spouse present), size of family, age of reference person, location of family (census region and MSA status), and race/ethnicity of the family’s reference person. Table 3.3 shows the number of families with family-level weights for each of the two panels separately, as well as the combined total and the total population estimate represented by the weighted total for all persons with person-level weights. Included as families in these counts are individuals living in one person RUs. There are 8,625 such families for Panel 6, Round 3 and 6,857 for Panel 7, Round 1. Thus, in total, there are 15,482 sample families in the file with positive family-level weights (WGTRU13>0). The population estimate of the number of these "family" units (families plus individual person units) with family-level weights containing at least one member of the U.S. civilian, non-institutionalized population is 121,753,042 based on summing the family level weights across all 15,482 MEPS families where WGTRU13 is positive. Table 3.3. Families with a family weight for the 2002 Point-in-Time file Panel 6 Panel 7 Combined Population estimate (weighted total of combined sample) Number 8,625 6,857 15,482 121,753,042 It should be noted that CPS and MEPS definitions of family units are slightly different. In particular, CPS does not include xxxxxx children in families or consider unmarried persons who live together as family units. Adjustments were made in the poststratification process to help compensate for some of these differences.

Assignment of Weights. If all key, in-scope members of a family responded to MEPS for their entire period of eligibility for Panel 67, Round 3 or for Panel 78, Round 1 and if the family had a key reference person, then that family received a family-level weight (WGTRU13>0). Reporting units consisting of an individual respondent who was both key and in-scope also received a family-level weight. These individual single person “family” units can be included or excluded from family-level analyses at the analyst’s discretion. Family-level weights were poststratified to figures obtained from the March 2002 2003 CPS. The family-level poststratification reflected reflects population distributions across family type (reference person married, spouse present; male reference person, no spouse present; female reference person, no spouse present), size of family, age of reference person, location of family (census region and MSA status), and race/ethnicity of the family’s reference person. Table 3.3 shows the number of families with family-level weights for each of the two panels separately, as well as the combined total and the total population estimate represented by the weighted total for all persons with person-level weights. Included as families in these counts are individuals living in one person RUs. There are 8,625 6,403 such families for Panel 67, Round 3 and 6,857 7,023 for Panel 78, Round 1. Thus, in total, there are 15,482 13,426 sample families in the file with positive family-level weights (WGTRU13>0). The population estimate of the number of these "family" units (families plus individual single person “family” units) with family-level weights containing at least one member of the U.S. civilian, non-institutionalized population is 121,753,042 123,860,517 based on summing the family level weights across all 15,482 13,426 MEPS families where WGTRU13 is positive. Table 3.3. Families with a family weight for the 2002 2003 Point-in-Time file Panel 6 7 Panel 7 8 Combined Population estimate (weighted total of combined sample) Number 8,625 6,857 15,482 121,753,042 6,403 7,023 13,426 123,860,517 It should be noted that CPS and MEPS definitions of family units are slightly different. In particular, CPS does not include xxxxxx children in families or consider unmarried persons who live together as family units. Adjustments were made in the poststratification process to help compensate for some of these differences.

Assignment of Weights. If all key, in-scope members of a family responded to MEPS for their entire period of eligibility for Panel 616, Round 3 or for Panel 717, Round 1 and if the family had a key reference person, then that family received a family-level weight (WGTRU13>0). Reporting units consisting of an individual respondent who was both key and in-scope also received a family-level weight. These individual single person “family” units can be included or excluded from family-level analyses at the analyst’s discretion. Family-level weights were poststratified to figures obtained from the March 2002 2012 CPS. The family-level poststratification reflected reflects population distributions across family type (reference person married, spouse present; male reference person, no spouse present; female reference person, no spouse present), size of family, age of reference person, location of family (census region and MSA status), and race/ethnicity of the family’s reference person. Table 3.3 3-3 shows the number of families with family-level weights for each of the two panels separately, as well as the combined total and the total population estimate represented by the weighted total for all persons families with personfamily-level weights. Included as families in these counts are individuals living in one person RUs. There are 8,625 7,671 such families for Panel 616, Round 3 and 6,857 7,919 for Panel 717, Round 1. Thus, in total, there are 15,482 15,590 sample families in the file with positive family-level weights (WGTRU13>0). The population estimate of the number of these "“family" ” units (families plus individual single person “family” units) with family-level weights containing at least one member of the U.S. civilian, non-institutionalized population is 121,753,042 135,994,714 based on summing the family level weights across all 15,482 15,590 MEPS families where WGTRU13 is positive. Table 3.33-3. Families with a family weight for the 2002 2012 Point-in-Time file Panel 6 16 Panel 7 17 Combined Population estimate (weighted total of combined sample) Number 8,625 6,857 15,482 121,753,042 7,671 7,919 15,590 135,994,714 It should be noted that CPS and MEPS definitions of family units are slightly different. In particular, CPS does not include xxxxxx children in families or consider unmarried persons who live together as family units. Adjustments were made in the poststratification process to help compensate for some of these differences. Again, note that MEPS population estimates have undergone some “discontinuities” as discussed earlier, as they now reflect 2010 Census data on population distributions.

Assignment of Weights. If all key, in-scope members of a family responded to MEPS for their entire period of eligibility for Panel 615, Round 3 or for Panel 716, Round 1 and if the family had a key reference person, then that family received a family-level weight (WGTRU13>0). Reporting units consisting of an individual respondent who was both key and in-scope also received a family-level weight. These individual single person “family” units can be included or excluded from family-level analyses at the analyst’s discretion. Family-level weights were poststratified to figures obtained from the March 2002 2011 CPS. The family-level poststratification reflected reflects population distributions across family type (reference person married, spouse present; male reference person, no spouse present; female reference person, no spouse present), size of family, age of reference person, location of family (census region and MSA status), and race/ethnicity of the family’s reference person. Table 3.3 3-3 shows the number of families with family-level weights for each of the two panels separately, as well as the combined total and the total population estimate represented by the weighted total for all persons families with personfamily-level weights. Included as families in these counts are individuals living in one person RUs. There are 8,625 5,967 such families for Panel 615, Round 3 and 6,857 8,149 for Panel 716, Round 1. Thus, in total, there are 15,482 14,116 sample families in the file with positive family-level weights (WGTRU13>0). The population estimate of the number of these "“family" ” units (families plus individual single person “family” units) with family-level weights containing at least one member of the U.S. civilian, non-institutionalized population is 121,753,042 133,803,893 based on summing the family level weights across all 15,482 14,116 MEPS families where WGTRU13 is positive. Table 3.33-3. Families with a family weight for the 2002 2011 Point-in-Time file Panel 6 15 Panel 7 16 Combined Population estimate (weighted total of combined sample) Number 8,625 6,857 15,482 121,753,042 5,967 8,149 14,116 133,803,893 It should be noted that CPS and MEPS definitions of family units are slightly different. In particular, CPS does not include xxxxxx children in families or consider unmarried persons who live together as family units. Adjustments were made in the poststratification process to help compensate for some of these differences. Again, note that MEPS population estimates have undergone some “discontinuities” due to an adjustment in the 2003 CPS estimates.

Assignment of Weights. If all key, in-scope members of a family responded to MEPS for their entire period of eligibility for Panel 613, Round 3 or for Panel 714, Round 1 and if the family had a key reference person, then that family received a family-level weight (WGTRU13>0). Reporting units consisting of an individual respondent who was both key and in-scope also received a family-level weight. These individual single person “family” units can be included or excluded from family-level analyses at the analyst’s discretion. Family-level weights were poststratified to figures obtained from the March 2002 2009 CPS. The family-level poststratification reflected reflects population distributions across family type (reference person married, spouse present; male reference person, no spouse present; female reference person, no spouse present), size of family, age of reference person, location of family (census region and MSA status), and race/ethnicity of the family’s reference person. Table 3.3 3-3 shows the number of families with family-level weights for each of the two panels separately, as well as the combined total and the total population estimate represented by the weighted total for all persons families with personfamily-level weights. Included as families in these counts are individuals living in one person RUs. There are 8,625 7,346 such families for Panel 613, Round 3 and 6,857 7,488 for Panel 714, Round 1. Thus, in total, there are 15,482 14,834 sample families in the file with positive family-level weights (WGTRU13>0). The population estimate of the number of these "“family" ” units (families plus individual single person “family” units) with family-level weights containing at least one member of the U.S. civilian, non-institutionalized population is 121,753,042 131,223,210 based on summing the family level weights across all 15,482 14,834 MEPS families where WGTRU13 is positive. Table 3.33-3. Families with a family weight for the 2002 2009 Point-in-Time file Panel 6 13 Panel 7 14 Combined Population estimate (weighted total of combined sample) Number 8,625 6,857 15,482 121,753,042 7,346 7,488 14,834 131,223,210 It should be noted that CPS and MEPS definitions of family units are slightly different. In particular, CPS does not include xxxxxx children in families or consider unmarried persons who live together as family units. Adjustments were made in the poststratification process to help compensate for some of these differences. Again, note that MEPS population estimates have undergone some “discontinuities” due to an adjustment in the 2003 CPS estimates.

Assignment of Weights. If all key, in-scope members of a family responded to MEPS for their entire period of eligibility for Panel 612, Round 3 or for Panel 713, Round 1 and if the family had a key reference person, then that family received a family-level weight (WGTRU13>0). Reporting units consisting of an individual respondent who was both key and in-scope also received a family-level weight. These individual single person “family” units can be included or excluded from family-level analyses at the analyst’s discretion. Family-level weights were poststratified to figures obtained from the March 2002 2008 CPS. The family-level poststratification reflected reflects population distributions across family type (reference person married, spouse present; male reference person, no spouse present; female reference person, no spouse present), size of family, age of reference person, location of family (census region and MSA status), and race/ethnicity of the family’s reference person. Table 3.3 3-3 shows the number of families with family-level weights for each of the two panels separately, as well as the combined total and the total population estimate represented by the weighted total for all persons families with personfamily-level weights. Included as families in these counts are individuals living in one person RUs. There are 8,625 5,136 such families for Panel 612, Round 3 and 6,857 7,821 for Panel 713, Round 1. Thus, in total, there are 15,482 12,957 sample families in the file with positive family-level weights (WGTRU13>0). The population estimate of the number of these "“family" ” units (families plus individual single person “family” units) with family-level weights containing at least one member of the U.S. civilian, non-institutionalized population is 121,753,042 130,610,298 based on summing the family level weights across all 15,482 12,957 MEPS families where WGTRU13 is positive. Table 3.33-3. Families with a family weight for the 2002 2008 Point-in-Time file Panel 6 12 Panel 7 13 Combined Population estimate (weighted total of combined sample) Number 8,625 6,857 15,482 121,753,042 5,136 7,821 12,957 130,610,298 It should be noted that CPS and MEPS definitions of family units are slightly different. In particular, CPS does not include xxxxxx children in families or consider unmarried persons who live together as family units. Adjustments were made in the poststratification process to help compensate for some of these differences. Again, note that MEPS population estimates have undergone some “discontinuities” due to an adjustment in the 2003 CPS estimates.

Assignment of Weights. If all key, in-scope members of a family responded to MEPS for their entire period of eligibility for Panel 618, Round 3 or for Panel 719, Round 1 and if the family had a key reference person, then that family received a family-level weight (WGTRU13>0). Reporting units consisting of an individual respondent who was both key and in-scope also received a family-level weight. These individual single person “family” units can be included or excluded from family-level analyses at the analyst’s discretion. Family-level weights were poststratified to figures obtained from the March 2002 2014 CPS. The family-level poststratification reflected reflects population distributions across family type (reference person married, spouse present; male reference person, no spouse present; female reference person, no spouse present), size of family, age of reference person, location of family (census region and MSA status), and race/ethnicity of the family’s reference person. Table 3.3 3-3 shows the number of families with family-level weights for each of the two panels separately, as well as the combined total and the total population estimate represented by the weighted total for all persons families with personfamily-level weights. Included as families in these counts are individuals living in one person RUs. There are 8,625 6,867 such families for Panel 618, Round 3 and 6,857 7,290 for Panel 719, Round 1. Thus, in total, there are 15,482 14,157 sample families in the file with positive family-level weights (WGTRU13>0). The population estimate of the number of these "“family" ” units (families plus individual single person “family” units) with family-level weights containing at least one member of the U.S. civilian, non-institutionalized population is 121,753,042 138,710,926 based on summing the family level weights across all 15,482 4,157 MEPS families where WGTRU13 is positive. Table 3.33-3. Families with a family weight for the 2002 2014 Point-in-Time file Panel 6 18 Panel 7 19 Combined Population estimate (weighted total of combined sample) Number 8,625 6,857 15,482 121,753,042 6,867 7,290 14,157 138,710,926 It should be noted that CPS and MEPS definitions of family units are slightly different. In particular, CPS does not include xxxxxx children in families or consider unmarried persons who live together as family units. Adjustments were made in the poststratification process to help compensate for some of these differences. Again, note that MEPS population estimates have undergone some “discontinuities” as discussed earlier, as they now reflect 2010 Census data on population distributions.

Assignment of Weights. If all key, in-scope members of a family responded to MEPS for their entire period of eligibility for Panel 68, Round 3 or for Panel 79, Round 1 and if the family had a key reference person, then that family received a family-level weight (WGTRU13>0). Reporting units consisting of an individual respondent who was both key and in-scope also received a family-level weight. These individual single person “family” units can be included or excluded from family-level analyses at the analyst’s discretion. Family-level weights were poststratified to figures obtained from the March 2002 2004 CPS. The family-level poststratification reflected reflects population distributions across family type (reference person married, spouse present; male reference person, no spouse present; female reference person, no spouse present), size of family, age of reference person, location of family (census region and MSA status), and race/ethnicity of the family’s reference person. Table 3.3 shows the number of families with family-level weights for each of the two panels separately, as well as the combined total and the total population estimate represented by the weighted total for all persons with person-level weights. Included as families in these counts are individuals living in one person RUs. There are 8,625 6,600 such families for Panel 68, Round 3 and 6,857 7,064 for Panel 79, Round 1. Thus, in total, there are 15,482 13,664 sample families in the file with positive family-level weights (WGTRU13>0). The population estimate of the number of these "family" units (families plus individual single person “family” units) with family-level weights containing at least one member of the U.S. civilian, non-institutionalized population is 121,753,042 124,832,233 based on summing the family level weights across all 15,482 13,664 MEPS families where WGTRU13 is positive. Table 3.3. Families with a family weight for the 2002 2004 Point-in-Time file Panel 6 8 Panel 7 9 Combined Population estimate (weighted total of combined sample) Number 8,625 6,857 15,482 121,753,042 6,600 7,064 13,664 124,832,233 It should be noted that CPS and MEPS definitions of family units are slightly different. In particular, CPS does not include xxxxxx children in families or consider unmarried persons who live together as family units. Adjustments were made in the poststratification process to help compensate for some of these differences.

Assignment of Weights. If all key, in-scope members of a family responded to MEPS for their entire period of eligibility for Panel 610, Round 3 or for Panel 711, Round 1 and if the family had a key reference person, then that family received a family-level weight (WGTRU13>0). Reporting units consisting of an individual respondent who was both key and in-scope also received a family-level weight. These individual single person “family” units can be included or excluded from family-level analyses at the analyst’s discretion. Family-level weights were poststratified to figures obtained from the March 2002 2006 CPS. The family-level poststratification reflected reflects population distributions across family type (reference person married, spouse present; male reference person, no spouse present; female reference person, no spouse present), size of family, age of reference person, location of family (census region and MSA status), and race/ethnicity of the family’s reference person. Table 3.3 shows the number of families with family-level weights for each of the two panels separately, as well as the combined total and the total population estimate represented by the weighted total for all persons with person-level weights. Included as families in these counts are individuals living in one person RUs. There are 8,625 6,405 such families for Panel 610, Round 3 and 6,857 7,417 for Panel 711, Round 1. Thus, in total, there are 15,482 13,822 sample families in the file with positive family-level weights (WGTRU13>0). The population estimate of the number of these "family" units (families plus individual single person “family” units) with family-level weights containing at least one member of the U.S. civilian, non-institutionalized population is 121,753,042 128,088,390 based on summing the family level weights across all 15,482 13,822 MEPS families where WGTRU13 is positive. Table 3.3. 3.3 Families with a family weight for the 2002 2006 Point-in-Time file Panel 6 10 Panel 7 11 Combined Population estimate (weighted total of combined sample) Number 8,625 6,857 15,482 121,753,042 6,405 7,417 13,822 128,088,390 It should be noted that CPS and MEPS definitions of family units are slightly different. In particular, CPS does not include xxxxxx children in families or consider unmarried persons who live together as family units. Adjustments were made in the poststratification process to help compensate for some of these differences. Again, note that MEPS population estimates have undergone some “discontinuities” recently due to an adjustment in the 2003 CPS estimates.

Assignment of Weights. If all key, in-scope members of a family responded to MEPS for their entire period of eligibility for Panel 619, Round 3 or for Panel 720, Round 1 and if the family had a key reference person, then that family received a family-level weight (WGTRU13>0). Reporting units consisting of an individual respondent who was both key and in-scope also received a family-level weight. These individual single person “family” units can be included or excluded from family-level analyses at the analyst’s discretion. Family-level weights were poststratified to figures obtained from the March 2002 2015 CPS. The family-level poststratification reflected reflects population distributions across family type (reference person married, spouse present; male reference person, no spouse present; female reference person, no spouse present), size of family, age of reference person, location of family (census region and MSA status), and race/ethnicity of the family’s reference person. Table 3.3 3-3 shows the number of families with family-level weights for each of the two panels separately, as well as the combined total and the total population estimate represented by the weighted total for all persons families with personfamily-level weights. Included as families in these counts are individuals living in one person RUs. There are 8,625 6,648 such families for Panel 619, Round 3 and 6,857 8,069 for Panel 720, Round 1. Thus, in total, there are 15,482 14,717 sample families in the file with positive family-level weights (WGTRU13>0). The population estimate of the number of these "“family" ” units (families plus individual single person “family” units) with family-level weights containing at least one member of the U.S. civilian, non-institutionalized population is 121,753,042 140,602,816 based on summing the family level weights across all 15,482 14,717 MEPS families where WGTRU13 is positive. Table 3.33-3. Families with a family weight for the 2002 2015 Point-in-Time file Panel 6 19 Panel 7 20 Combined Population estimate (weighted total of combined sample) Number 8,625 6,857 15,482 121,753,042 6,648 8,069 14,717 140,602,816 It should be noted that CPS and MEPS definitions of family units are slightly different. In particular, CPS does not include xxxxxx children in families or consider unmarried persons who live together as family units. Adjustments were made in the poststratification process to help compensate for some of these differences. Again, note that MEPS population estimates have undergone some “discontinuities” as discussed earlier, as they now reflect 2010 Census data on population distributions.

Assignment of Weights. If all key, in-scope members of a family responded to MEPS for their entire period of eligibility for Panel 614, Round 3 or for Panel 715, Round 1 and if the family had a key reference person, then that family received a family-level weight (WGTRU13>0). Reporting units consisting of an individual respondent who was both key and in-scope also received a family-level weight. These individual single person “family” units can be included or excluded from family-level analyses at the analyst’s discretion. Family-level weights were poststratified to figures obtained from the March 2002 2010 CPS. The family-level poststratification reflected reflects population distributions across family type (reference person married, spouse present; male reference person, no spouse present; female reference person, no spouse present), size of family, age of reference person, location of family (census region and MSA status), and race/ethnicity of the family’s reference person. Table 3.3 3-3 shows the number of families with family-level weights for each of the two panels separately, as well as the combined total and the total population estimate represented by the weighted total for all persons families with personfamily-level weights. Included as families in these counts are individuals living in one person RUs. There are 8,625 6,715 such families for Panel 614, Round 3 and 6,857 6,644 for Panel 715, Round 1. Thus, in total, there are 15,482 13,359 sample families in the file with positive family-level weights (WGTRU13>0). The population estimate of the number of these "“family" ” units (families plus individual single person “family” units) with family-level weights containing at least one member of the U.S. civilian, non-institutionalized population is 121,753,042 132,865,347 based on summing the family level weights across all 15,482 13,359 MEPS families where WGTRU13 is positive. Table 3.33-3. Families with a family weight for the 2002 2010 Point-in-Time file Panel 6 14 Panel 7 15 Combined Population estimate (weighted total of combined sample) Number 8,625 6,857 15,482 121,753,042 6,715 6,644 13,359 132,865,347 It should be noted that CPS and MEPS definitions of family units are slightly different. In particular, CPS does not include xxxxxx children in families or consider unmarried persons who live together as family units. Adjustments were made in the poststratification process to help compensate for some of these differences. Again, note that MEPS population estimates have undergone some “discontinuities” due to an adjustment in the 2003 CPS estimates.

Assignment of Weights. If all key, in-scope members of a family responded to MEPS for their entire period of eligibility for Panel 617, Round 3 or for Panel 718, Round 1 and if the family had a key reference person, then that family received a family-level weight (WGTRU13>0). Reporting units consisting of an individual respondent who was both key and in-scope also received a family-level weight. These individual single person “family” units can be included or excluded from family-level analyses at the analyst’s discretion. Family-level weights were poststratified to figures obtained from the March 2002 2013 CPS. The family-level poststratification reflected reflects population distributions across family type (reference person married, spouse present; male reference person, no spouse present; female reference person, no spouse present), size of family, age of reference person, location of family (census region and MSA status), and race/ethnicity of the family’s reference person. Table 3.3 3-3 shows the number of families with family-level weights for each of the two panels separately, as well as the combined total and the total population estimate represented by the weighted total for all persons families with personfamily-level weights. Included as families in these counts are individuals living in one person RUs. There are 8,625 7,314 such families for Panel 617, Round 3 and 6,857 7,508 for Panel 718, Round 1. Thus, in total, there are 15,482 14,822 sample families in the file with positive family-level weights (WGTRU13>0). The population estimate of the number of these "“family" ” units (families plus individual single person “family” units) with family-level weights containing at least one member of the U.S. civilian, non-institutionalized population is 121,753,042 138,154,576 based on summing the family level weights across all 15,482 14,822 MEPS families where WGTRU13 is positive. Table 3.33-3. Families with a family weight for the 2002 2013 Point-in-Time file Panel 6 17 Panel 7 18 Combined Population estimate (weighted total of combined sample) Number 8,625 6,857 15,482 121,753,042 7,314 7,508 14,822 138,154,576 It should be noted that CPS and MEPS definitions of family units are slightly different. In particular, CPS does not include xxxxxx children in families or consider unmarried persons who live together as family units. Adjustments were made in the poststratification process to help compensate for some of these differences. Again, note that MEPS population estimates have undergone some “discontinuities” as discussed earlier, as they now reflect 2010 Census data on population distributions.

Assignment of Weights. If all key, in-scope members of a family responded to MEPS for their entire period of eligibility for Panel 611, Round 3 or for Panel 712, Round 1 and if the family had a key reference person, then that family received a family-level weight (WGTRU13>0). Reporting units consisting of an individual respondent who was both key and in-scope also received a family-level weight. These individual single person “family” units can be included or excluded from family-level analyses at the analyst’s discretion. Family-level weights were poststratified to figures obtained from the March 2002 2007 CPS. The family-level poststratification reflected reflects population distributions across family type (reference person married, spouse present; male reference person, no spouse present; female reference person, no spouse present), size of family, age of reference person, location of family (census region and MSA status), and race/ethnicity of the family’s reference person. Table 3.3 shows the number of families with family-level weights for each of the two panels separately, as well as the combined total and the total population estimate represented by the weighted total for all persons with person-level weights. Included as families in these counts are individuals living in one person RUs. There are 8,625 6,680 such families for Panel 611, Round 3 and 6,857 5,736 for Panel 712, Round 1. Thus, in total, there are 15,482 12,416 sample families in the file with positive family-level weights (WGTRU13>0). The population estimate of the number of these "family" units (families plus individual single person “family” units) with family-level weights containing at least one member of the U.S. civilian, non-institutionalized population is 121,753,042 129,215,973.78 based on summing the family level weights across all 15,482 12,416 MEPS families where WGTRU13 is positive. Table 3.3. 3.3 Families with a family weight for the 2002 2007 Point-in-Time file Panel 6 11 Panel 7 12 Combined Population estimate (weighted total of combined sample) Number 8,625 6,857 15,482 121,753,042 6,680 5,736 12,416 129,215,974 It should be noted that CPS and MEPS definitions of family units are slightly different. In particular, CPS does not include xxxxxx children in families or consider unmarried persons who live together as family units. Adjustments were made in the poststratification process to help compensate for some of these differences. Again, note that MEPS population estimates have undergone some “discontinuities” recently due to an adjustment in the 2003 CPS estimates.