RESEARCH PROPOSALS

A research proposal consists of a research question, research hypothesis, a previous study that justifies your hypothesis in terms of LAST, sample materials, design, procedure, and predicted results

• The research question asks about some unusual phenomenon or property of language. For example,

How does the parser use event structure?

• The research hypothesis provides a possible answer to the research question. For example,

If a telic verb elicits a slot for a delimiter, then reading times for direct objects will be faster for telic verbs than for atelic verbs. This effect will be greater for specific objects than for non-specific objects.

The research hypothesis is an if-then statement in which the if-part suggests a

mechanism.

The proposed mechanism requires that verbs are represented in the mental lexicon as telic vs. atelic. This information becomes available as soon as the verb is recognized. Furthermore, telic events require something that delimits the event, marking its end, and the parser uses this fact to form an initial meaning hypothesis. Telic verbs are more likely to have a direct object (or patient) that undergoes a change of state.

The then-part of the research hypothesis describes how this mechanism can be

observed in behavior.

The research hypothesis proposes that this mechanism can be tested by observing reading times for direct objects. The prediction that follows from the proposed mechanism is that telic verbs cause faster reading times for direct objects, compared to atelic verbs. The basic independent variable manipulation is verb telicity.

• A research proposal contains a description of (at least) one previous study that justifies the hypothesis may be a linguistic analysis of some interesting property, such as event structure, sentence structure, etc., or an experimental study that examines behavioral measures of sentence comprehension

Candidates for a proposal on event structure might be Vendler, Seegmiller et al., Magliano & Schleich, etc. The proposal describes the basic linguistic analysis with examples and (if an experiment) it describes the materials (with examples), procedure, results in qualitative terms, and theoretical implications (especially with regard to your hypothesis).

• Sample materials show one complete set of materials that illustrate all design conditions that you would use in a quantitative behavioral study of sentence comprehension. Your description of sample materials also states the number of critical sentence sets and the number of filler sentences and their properties.

There are 16 critical sentence sets and 64 filler sentences that present a wide range of sentence structures.

Specific Objects:

Potentially Intransitive:

Atelic verb: The man rested the player.

Telic verb: The man tripped the player.

Transitive Only:

Atelic verb: The man admired the player.

Telic verb: The man captured the player.

Non-Specific Objects:

Potentially Intransitive:

Atelic verb: The man rested players.

Telic verb: The man tripped players.

Transitive Only:

Atelic verb: The man admired players.

Telic verb: The man captured players.

• Design states the independent variables, their levels, how they are manipulated in an experiment (i.e., within or between participants), and how many lists are in the experiment.

The experiment has a 2x2x2 design with both variables manipulated within participants: Sub-categorization (potentially intransitive/transitive only) x Verb Telicity (telic/atelic) x Object Specificity (specific/non-specific). There are 8 lists of materials.

• Procedure describes the participant’s task. What does the participant see or hear, what does the participant have to do, and what aspect of behavior do you measure?

Participants read sentences one word at a time in a self-paced moving window reading task. After each sentence the screen presents a comprehension question about the content of the sentence. Participants answer the question by pressing the right shift key (YES) or the left shift key (NO). These answers and response times are recorded. The primary data are reading times for words in the direct object of the critical sentences.

• A research proposal describes the predicted results in terms of a graph that shows in qualitative terms the behavioral results that your research hypothesis predicts.

For the hypothesis about event structure, the predicted results would show two graphs with reading times for direct objects on the vertical axis. One graph would show the predicted results for PI verbs and the other those for TO verbs. Each graph would have the levels of Verb Telicity on the horizontal axis and two lines in the body of the graph, one for specific objects and the other for non-specific objects. The lines in the body of the graph would show that for telic verbs reading times are faster for specific objects than for non-specific objects, for both PI and TO verbs, and that for atelic verbs reading times do not differ for specific and non-specific objects. In addition, the reading times for direct objects will be faster overall for telic verbs than for atelic verbs.