影響唸詞速度的是詞頻而非習得年齡：童年詞頻的角色

Word Frequency, not Age of Acquisition, Affects Word Naming: The Role of Childhood Frequency

本論文以兩個實驗，驗證習得年齡（AoA）與詞頻對唸詞速度的影響。使用的詞頻，除了成人詞頻以外，還包括計算自國小課本出現次數的兒童詞頻。不同於一般利用主觀評量所獲得的AoA，本實驗的AoA為該詞實際出現在國小課本年級的真正習得年齡。實驗一是以成年詞頻作為分組與配對依據，其中早習得－低頻詞有兩組材料，一為低兒童詞頻（CL），一為高兒童詞頻（CH）。結果顯示：CL材料組的AoA與詞頻都有效果，但無交互作用；CH材料組，AoA與詞頻都有效果，並有顯著的交互作用；在低頻組的AoA效果，CH組比CL組大許多（36 vs. 19 ms）。實驗二改以兒童詞頻和成人詞頻之和，作為分組與配對依據。結果顯示，不論兒童詞頻與成人詞頻分配是採單純或混合方式，兩組材料的AoA效果都消失不見，只剩下顯著的詞頻效果。本研究結果意涵，過去相關研究常見的AoA效果，有可能是未適當控制兒童詞頻，為兒童詞頻效果的假象。

Two experiments were conducted to assess the importance of age-of-acquisition (AoA) and frequency in a word naming task. In addition to adult- hood- frequency, childhood- frequency calculated from elementary textbooks of participants was used. A real age- of- acquisition was taken to be the age at which the word was first read on the textbook. In Experiment 1, the traditional adulthood- frequency was used for grouping and matching and two different word sets of early-acquired, low-frequency was chosen by their childhood- frequency to be low (CL) or high (CH). With CL words, both AoA and frequency effects had independent effects but no effect of interaction. With CH words, both AoA and frequency effects had independent effects and a significant interaction effect was also found. The AoA effect in low frequency was actually larger for CH words than for CL words (36 vs. 19 ms). In Experiment 2, the sum of childhood- frequency and adulthood-frequency was used for grouping and matching. A significant frequency effect was found but no AoA effect on either mixed or pure word sets once the total frequency was controlled. These results implicated that previous reports of effect of AoA on word- naming speed might be artifactual, being due to a confounding of AoA with childhood- frequency.