we can generalize to a population consisting of whoever responded. Whether this is representative for any general population is beyond statistical inference. 1.4.3 Response and Nonresponse Error Nonresponse is the inability to obtain data for all sampled units on all questions. There are two types of nonresponse in surveys: unit nonresponse and item nonresponse. Unit nonresponse is the failure to obtain any information from an eligible sample unit. Unit nonresponse can be the result of noncontact or refusal. Lynn (Chapter 3) provides an extensive overview on nonresponse and nonresponse error; for a discussion of nonresponse error in cross-cultural studies, see Couper and 13 de Leeuw (2003); for statistical adjustment and weighting see Biemer and Christ (Chapter 16). Item-nonresponse or item missing data refers to the failure to obtain information for one or more
multiple aspects of packaging, including high oxygen concentrations in MAP in order recyclability, simple packaging, reusable, to promote color stability (Zakrys et al. refillable, renewable materials, less materi- 2008). The color of lamb may also be als, less or no plastics, and bulk rather than extended by storage under MAP conditions individual packaging (Doyle 2008). Also, (Kerry et al. 2000). the noncontact preservative effect of active MAP has now been available to producers packaging offers the opportunity to produc- for many years. As far back as 1933, Killefer ers of prolonging shelf life further, while (1930), using 100% carbon dioxide (CO2) at maintaining the clean label status of meat 4–7°C, found that pork and lamb remained products. fresh for twice as long as equivalent products Figure 13.1
annaabi.ee 
