从历史发展看，描述统计为前驱，推断统计为核心，实验设计为后衍；从研究过程看，应先做研究设计，再做描述统计，最后做推断统计。

( a ) Var a, b, c, d, e, f, g, h, i; semaphore:= 0, 0, 0, 0, 0, 0, 0, 0, 0; begin parbegin begin S1; signal(a); signal(b); end; begin wait(a); S2; signal(c); signal(d); end; begin wait(b); S3; signal(e); signal(f); end; begin wait(c); S4; signal(g); end; begin wait(d); wait(e); S5; signal(h); end; begin wait(f); S6; signal(i); end; begin wait(g); wait(h); wait(i); S7; end; parend end ( b ) Var a, b, c, d, e, f, g, h, i, j, k, l; semaphore:= 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0; begin parbegin begin S1; signal(a); signal(b); end; begin wait(a); S2; signal(c); signal(d); signal(e); end; begin wait(b); S3; signal(f); signal(g); signal(h); end; begin wait(c); S4; signal(i); end; begin wait(d); wait(f); S5; signal(j); end; begin wait(e); wait(g); S6; signal(k); end; begin wait(h); S7; signal(l); end; begin wait(i); wait(j); wait(k);wait(l); S8; end; parend end