No abstract

Sociometry is concerned with networks of relationship among groups of people. If the group is very large, the work of tracing all the relationships becomes tedious, and the task of describing the resulting net precisely becomes impossible. Here the problem of such large sociometric nets is approached with probabilistic and statistical methods.

Concurrent hemolysis in patients with vitamin B12 deficiency is a well-recognized phenomenon and has been attributed to intramedullary destruction of erythrocytes (ineffective erythropoiesis). Recent studies revealed that homocysteine increased the risk of hemolysis in vitamin B12 deficiency in vitro and there is a high frequency (30%) of vitamin B12 deficiency in asymptomatic patients with homozygous methylene tetrahydrofolate reductase (MTHFR) C677T mutation, a known cause of hyperhomocysteinemia. Here we report three patients with MTHFR mutations and vitamin B12 deficiency presenting with hemolytic anemia and severely elevated homocysteine levels. Patients demonstrated complete resolution of hemolysis with simultaneous normalization of serum homocysteine levels after vitamin B12 treatments. We reviewed pertinent literature, and hypothesized that hemolytic anemia may be more prevalent in patients who have a coexisting MTHFR gene mutation and vitamin B12 deficiency possibly related to severely elevated homocysteine levels. The hemolysis in these cases occurred predominantly in peripheral blood likely due to the combined effects of structurally defective erythrocytes and homocysteine-induced endothelial damage with microangiopathy.

The frequency of verbal interactions among members of small discussion groups has been carefully studied in experiments by Bales and his associates at Harvard and by Stephan and Mishler at Princeton. Stephan suggested that the data could be described by a simple exponential equation, which has intrigued several mathematical sociologists because of its apparent generality in providing a concise statement of experimental results. The present paper suggests that this result could be generated by a simple probabilistic process, which may have uses in other types of group interaction studies where an analogous structure prevails.

No abstract