Cross-sectional research based on NSFH data found that family process variables appeared to most significantly influence the quality of parent-child relationship (Lansford, Ceballo, Abbey, & Stewart, 2001). Reports on relationships between family members, overall family atmosphere, and parent and child well-being were collected from 799 families representing five different family structures (intact adoptive parents (14%), intact biological parents (25%), divorced single mother (25%), biological mother and stepfather (25%), and biological father and stepmother (11%)). Analysis of the data showed that mothers in stepmother and stepfather families reported less frequent arguments than mothers in intact biological families or adoptive families, and that adoptive mothers and two-parent biological fathers reported higher "family cohesion", measured in reports of fun, compassion, love, and teamwork versus tenseness, stress, and distance in the family, than other mothers and fathers, respectively. The researchers concluded that family processes, such as interpersonal disagreements, were more accurate predictors of family relationships than family structure was found to be. Similarly, research conducted by Musick and Bumpass (1998), based on a subsample of NSFH data, suggests that marital discord, in general, is associated with expressions of less positive parent-child relationships. The participants consisted of 842 non-Hispanic Caucasian children ages 12 to 18, and their mothers and/or fathers. Cross-sectional data showed that parents in stepfamilies were less likely to show affection to their children, in hugs or praise, than their intact marriage or single counterparts. The researchers found that parents from medium- or high-conflict intact families are more likely to hit or yell at their children than parents from single parent families, step-parent families, or low-conflict intact families.